Embed Size (px)
Citation preview
Master’s thesis
Technology Acceptance Model: A Case of Electronic Health Record in Estonia
Author: Veronika Losova Institution: M. Soc. Sc. Organizational Innovation and Entrepreneurship
Copenhagen Business School
Supervisor: Kirstine Zinck Pedersen, PhD, Assistant Professor Center for Health Management, Department of Organization, CBS
Master’s thesis: STUs: 151 569 Pages: 71 Submission date: 18th November 2014
ACKNOWLEDGEMENT
At this place, I would like to express my thanks to people who helped me materialize this master’s
thesis.
Firstly, I’d like to thank my thesis supervisor Kirstine Zinck Pedersen, PhD for taming my
“consultancy-like” ambitions and forming them into more academically appropriate analytical focus.
Secondly, I’d like to thank Dr. Peeter Ross for introducing me to the topic of Estonian EHR. I’d also
like to thank to all the interviewees that participated in this research.
Thirdly, I’d also like to thank to Petr Bilik, for providing valuable feedback towards the final stages of
thesis writing.
My thanks belong also to Danielle Astor, Alejandra P. Betancourt, Kristine Lorenzen and Andra Tolbus
for proofreading different chapters of this thesis.
Last, but not least, I like to thank my dad, Milan Losa, for continuous reminders that the thesis won’t
write itself.
ABSTRACT The goal of this master’s thesis is to understand key motivators and activities supporting high levels of
national Electronic Health Record (EHR) usage by healthcare professionals in Estonia and relate them
to findings from Technology Acceptance Model (TAM) research. This thesis aims at contributing both,
to better understanding of Estonian EHR implementation case, and TAM application in national public
healthcare settings.
In this thesis, I found out that the key motivator for the high levels of usage of the system was legal act
making it mandatory for healthcare professionals to send data to the national EHR. Despite that, I
discovered that TAM variables perceived usefulness, perceived ease of use and psychological ownership
played a role in Estonian EHR implementation case in positively influencing attitude towards the
system. In this thesis, I argue that attitude towards the system is an important predictor of success for
public systems, even if their usage is mandated by the law. I also argue that importance of psychological
ownership may be especially important in the mandatory use cases.
Furthermore, I discovered that in this particular case, information quality and job relevance influenced
perceived usefulness, and that compatibility, self-efficacy and support influenced perceived ease of use.
In Estonia, compatibility however had controversial effect on perceived ease of use and I argue that in
this case, it may have influenced it negatively rather than positively.
1
TABLE OF CONTENTS
1 Introduction ......................................................................................................................... 3
2 Methodology ....................................................................................................................... 5 2.1 Qualitative research ........................................................................................................... 5 2.2 Case study approach ......................................................................................................... 6
2.2.1 Case selection and scope .............................................................................................. 6 2.3 Research design ................................................................................................................. 6
2.3.1 Secondary data collection ............................................................................................... 7 2.3.2 Primary data collection ................................................................................................... 7
3 Technology Acceptance Model ...................................................................................... 11 3.1 Importance of Technology Acceptance Model .............................................................. 11 3.2 Original Technology Acceptance Model ........................................................................ 12
3.2.1 Underlying research ..................................................................................................... 13 3.3 Technology Acceptance Model extensions ................................................................... 14
3.3.1 TAM2 ............................................................................................................................ 14 3.3.2 Technology Acceptance Model extensions in healthcare setting ................................. 15
3.4 TAM theory application in this thesis ............................................................................. 22 3.4.1 Level of use vs. intention to use vs. attitude to a system .............................................. 23 3.4.2 TAM variables summary ............................................................................................... 24
4 Analysis – Estonian EHR description ............................................................................. 30 4.1 EHR in Estonian: system features, goals and vision .................................................... 30 4.2 Technical solution ............................................................................................................ 31 4.3 Estonian healthcare system ............................................................................................ 32 4.4 EHR - Institutional organization ...................................................................................... 32
4.4.1 Estonian e-Health Foundation ...................................................................................... 33 4.5 Timeline ............................................................................................................................. 34
5 Analysis – Adoption of Estonian EHR by healthcare professionals ........................... 37 5.1 Mandatory system usage and TAM ................................................................................. 37
5.1.1 What made it possible to put the mandatory usage into the law? ................................ 39
2
5.1.2 How does mandatory usage influences TAM? ............................................................. 40 5.2 Role of perceived usefulness in Estonian EHR implementation ................................. 44
5.2.1 Information availability critical for perceived usefulness ............................................... 44 5.2.2 Improving perceived usefulness by introducing e-services .......................................... 48
5.3 Role of perceived ease of use in Estonian EHR implementation ................................ 50 5.3.1 Compatibility with existing processes and systems ...................................................... 51 5.3.2 Effect of computer skills on perceived ease of use ....................................................... 55
5.4 Role of psychological ownership in Estonian EHR implementation ........................... 58 5.4.1 Involving healthcare professionals ............................................................................... 60 5.4.2 Role of EEHF in building psychological ownership ....................................................... 61
5.5 Summary: Elements influencing high attitude, acceptance or usage levels of
Estonian national EHR .............................................................................................................. 63
6 Conclusions ...................................................................................................................... 64
7 Discussion ........................................................................................................................ 68
8 Table of figures ................................................................................................................. 71
Bibliography ............................................................................................................................ 72
3
INTRODUCTION Currently, public health services in Europe face numerous challenges; from rising occurrence of
diseases such as diabetes, cancer or cardiovascular diseases, through significant difference in life
expectancy and healthcare opportunities both between different European countries and population
groups; to ageing population, increasing costs of care and restrictions in public funding. (Koppel,
Leventhal, & Sedgley, 2009) As a consequence, doctors are expected to deal with daunting patient loads
and expectations for short office visits. (McAlearney, Song, Jorina, Hirsch, Kowalczyk, & Chisolm,
2010)
Electronic health records (EHR) are deemed to be an important part of the solution to these challenges,
as they are expected to help improve quality, safety, efficiency and coordination of care (McAlearney,
Song, Jorina, Hirsch, Kowalczyk, & Chisolm, 2010), to improve the quality of patient care by
increasing care coordination, eliminating errors, and reducing costs (Moores, 2012), to improve patient
care and outcomes, increase efficiency, lower costs, improve billing processes, reduce of the frequency
of lost records, data and medication errors and improve access to patient histories (Emmanouilidou &
Burke, 2013), to enhance the safety, quality and efficiency of healthcare (Cresswell, Worth, & Sheikh,
2012) etc.
Therefore, many countries are now investing substantial sums of money in EHR system. (Cresswell,
Worth, & Sheikh, 2012) However, despite the long history of EHR, the progress on EHR deployment
remains below expectations. (Emmanouilidou & Burke, 2013) There seems to be numerous failure
examples across the countries, institutions and types of EHR (Ellingsen & Monteiro, 2012) (Cresswell,
Worth, & Sheikh, 2012) (Berg, 2001), as only handful of EHR implementation process were
successfully carried out. (Tavakoli, Jahanbakhsh, Mokhtari, & Tadayon, 2011)
One of the most common reasons for failure is rejection of the system by the users, as the system is not
perceived as being fit-for-purpose. (Cresswell, Worth, & Sheikh, 2012) McAlearney et al. confirm that
gaining physician buy-in has proved extremely difficult. (McAlearney, Song, Jorina, Hirsch,
Kowalczyk, & Chisolm, 2010) Moores observes that the introduction of health IT can often generate
resistance from healthcare workers, especially if the system is seen as a threat or incompatible with the
way they like to work. (Moores, 2012)
4
Yet, the improvements to healthcare mentioned above will only occur if clinicians access the key
functions and use them regularly. (Simon, et al., 2007) In addition, Simon et al. also note that the
benefits will only be realized when clinicians use “higher level” functions supporting advanced clinical
decision-making. (Simon, et al., 2007)
Estonia seems to be one of the rare examples of a country, which achieved to implement a national EHR
system that is used by 95 % of healthcare professionals and includes a medical record of more than 75
% of Estonian population. (Tiik, 2012) This seems to be an advanced level of usage compared to other
countries striving to implement such a system, as most of the countries had to abandon long-planned
initiatives (Great Britain and the Czech Republic), are at the beginning of the process (Australia, Austria
etc.) or deal with interoperability issues of different local or regionals solutions (Canada, Denmark etc.).
According to an OECD and the European Commission report on ICT implementation and adoption in
healthcare, Estonia took the top spot among 30 countries included in the study, ahead of Denmark and
Sweden. (Nortal, 2013) The Estonian Health Information System is a globally unique e-health solution
as it encompasses the whole country, registers virtually all residents’ medical history from birth to
death, and is based on the comprehensive state-developed basic IT infrastructure. (Tiik & Ross, 2010)
Estonia thus seems like a good case to study in order to understand dynamics that allowed this country
to achieve such a high usage levels of the system.
Aim of the master’s thesis In this master’s thesis, I am going to focus on uncovering reasons that lead to high usage levels of
Estonian EHR. I am going to answer questions such as, why Estonian healthcare professionals use the
system, what motivates their behavior and what activities were undertaken in order to support this
motivation. I am going to use Technology Acceptance Model (TAM) as an analytical framework
allowing me to describe and theoretically support different usage drivers. The goal is to understand key
motivators of the EHR usage by healthcare professionals that occurred in Estonia and to relate them to
findings from the TAM research. This thesis aims at contributing both, to better understanding of
Estonian EHR implementation case, and TAM application in national public healthcare settings.
5
METHODOLOGY This chapter of the master’s thesis describes the methods used to write the thesis. It includes reasoning
on type of research used, research design and analysis of this master’s thesis.
1.1 Qualitative research
This is a qualitative research study. Up until the mid-1990s, for a significant proportion of mainstream
IS scholars, scientific research meant a quantitative research, and studies not using the so-called
“scientific methods” were excluded from the definition of research itself. (Sarker, Xiao, & Beaulieu,
2013) The same applies to the TAM research. Most of the TAM studies found were quantitative studies
measuring effects of different TAM variables on user acceptance, or they were a qualitative meta-
analysis of the TAM research literature.
Measurements tell us how often or how many people behave in a certain way but they do not adequately
answer the question “why”. Research which attempts to increase our understanding of why things are
the way they are in our social world and why people act the ways they do is the qualitative research.
(Hancock, 1998)
As TAM in this master’s thesis is mainly used to answer a “why” question and to analyze human
behavior and the activities that affected the behavior, it seems to be more appropriate to use qualitative
research methods in this case. Qualitative approach also fits aim of this study better, as it is focused on
explaining certain aspects of the particular subject of study. (Cormack, 1991) It is also more appropriate
to use qualitative research method, as the aim is to gain detailed understanding about implementation of
national Electronic Health Record in Estonia, discover what the main usage motivators for healthcare
professionals were and analyze them using Technology Acceptance Model (TAM) and its extensions.
As the aim of the thesis is to obtain in depth understanding of the phenomenon, the qualitative approach
offers better toolset. It also allows flexibility and the attainment of deeper, more valid understanding of
the subject than could be achieved with more rigid approach (Duffy, 1986)
6
1.2 Case study approach
The case study approach is an empirical research method, which investigates a case in a given context. It
is desired to investigate the case in its richness and depth. (Eriksson & Kovalainen, 2008) This master’s
thesis is a single case study. It focuses on Estonian EHR implementation case. By describing and
analyzing this case in its depth, this master’s thesis should uncover complexities and contradictions
present in this case. (Flyvbjerg, 2006)
1.2.1 Case selection and scope
My first interaction with Estonian EHR happened in February 2013 at Vitalis, a healthcare IT
conference in Gothenburg, Sweden. English track of the conference comprised, among others, of four
lectures on national EHR systems in Finland, Denmark, Norway and Estonia. Dr. Peeter Ross was
presenting the Estonian system. From the presentations, it was apparent that among afore mentioned
four countries, Estonia is furthest ahead in implementing the system in terms of levels of system usage.
By selecting Estonian EHR implementation case for this master’s thesis, I intended to select an extreme
case – a case that would especially well describe a successful implementation of an EHR system.
(Flyvbjerg, 2006) Originally, I wanted to describe all the different factors facilitating a successful
implementation of the Estonian EHR using ANT theory. However, preliminary research showed that
this task would be too broad for the purposes of master’s thesis at Copenhagen Business School. Firstly,
the definition of “a successful implementation” is a tricky one and it would take a whole master’s thesis
to assess if and under which conditions it is possible to call the Estonian EHR implementation a
successful one. Secondly, the whole implementation process was a complex process, mapping of which
would require substantially more space and time. Therefore, after careful considerations, I chose to
focus on uncovering reasons for high usage levels of the system by Estonian healthcare professionals.
This seemed to be an appropriate delimitation, as most of the “success” discourse in Estonian EHR case
is based on the high usage levels.
1.3 Research design
The research for this master’s thesis uses a combination of primary and secondary data collection.
7
1.3.1 Secondary data collection
There were two types of secondary data used. Firstly, academic research focused on Technology
Acceptance Model (TAM) application on healthcare cases were used. Findings of these articles are
summarized in the theoretical chapter of this master’s thesis. These findings were used to structure and
analyze information obtained in primary data research to enrich the case narrative with relevant
theoretical background. Furthermore, reflections of the case on these findings were also commented on
in the analytical chapter and summarized in the conclusions.
Secondly, various reports and online data were used to inform description of Estonian EHR. These
served to gain understanding of the case by presenting official accounts of events as they occurred for
the author and the reader. Most of the findings drawn from these documents are presented in chapter 0.
Amount of data accessible on Estonian EHR in English was limited and they were mostly coming from
the official website of Estonian e-Health Foundation (EEHF).
1.3.2 Primary data collection
Semi-structured interview was the key method of collecting primary data. Semi-structured interview is
performed based on an interview guide, which contains key areas of interest for the interviewer to
follow, but gives enough space to venture to areas of interest of the interviewee, if the interviewer
considers to be interesting for the particular study.
This approach was appropriate for this study, as it enabled me to uncover specificities of this case. For
example, using structured interviews would only allow me to focus on predefined set of variables.
Seven oral interviews were conducted for the purpose of this master’s thesis between April and October
2014. Interviewees were people involved in the implementation process of the Estonian national EHR in
the role of administrators or policy bearers.
I have selected to focus on the administrators of the process as it was my believe that these people
should have the overview over the process and should provide rather aggregated account based on their
observations and experience, as compared to for example healthcare professionals, who would have
most probably provided a subjective account only based on their own feelings. The administrators were
primarily from EEHF, the organization responsible for the implementation of the system. One person
from the Ministry of Social Affaires, the political patron of EHR, and one from Estonian Health
Insurance Fund were interviewed as well based on suggestion of the EEHF interviewees.
8
First, I contacted Peeter Ross, whom I had met at Vitalis 2013 conference in Gothenburg, Sweden,
February 2013. Dr. Ross was for a long time an active member of EEHF’s board, where he represented
East Tallinn Central Hospital and he has published several articles on the Estonian EHR. I held my first
interview with him in April 2014. This interview served as an expert interview as well, uncovering EHR
system features. Following interviewees were approached based on Dr. Ross’s recommendations, as
well as based on list of employees of EEHF published on the EEHF website.
Overview of interviewees is in Figure 1. As there was a limited number of players in the whole process,
only limited number of information can be provided while keeping their anonymity. Reader should also
note that Dr. Ross is not interviewee 1 in this master’s thesis. .
No. Qualification / Education
Role in the process
Interviewee 1 Business EEHF employee, Ministry of Social Affaires Interviewee 2 IT EEHF employee Interviewee 3 MD Ministry of Social Affaires Interviewee 4 MD EEHF board, a hospital representative Interviewee 5 MD EEHF board, a healthcare providers representative Interviewee 6 IT Estonian Health Insurance Fund employee Interviewee 7 MD EEHF employee
Figure 1: Interviewees, source: author
Timing and topics First two interviews were performed in April 2014, another five interviews in September and early
October 2014. The later interviews were shorter; they took 45 minutes on average; compared to the
earlier ones that took more than 1 hour each. This was caused by narrowing the scope of the master’s
thesis during the research process as described above. Narrowing the scope down allowed me to leave
out general questions about the whole implementation process. In addition, the first interviews also
served as expert interviews, describing the features of the Estonian EHR. The interviews performed in
Autumn 2014 focused almost exclusively on the activities meant to promote usage among the
physicians.
Performing the interviews The interviews were semi-structured interviews. For each participant, I prepared adjusted set of areas to
cover. The set was informed by the role of the interviewee in the implementation process, as well as on
learnings gained from previous interviews. In order to give the interviewees an opportunity to warm up,
9
each interview started with question asking them to describe their role in the system. Subsequent
questions focused on the reasons each interviewees identified influencing levels of usage of the EHR by
healthcare professionals. Each interview was concluded with giving the interviewees a chance to add
whatever they thought they hadn’t had chance to say in the interviewee. None of them added anything.
Interviews were performed by the author of this thesis.
The interviews were performed online, via Skype technology. For most of the interviews, camera
couldn’t have been used in order not to disturb quality of the sound by overloading the internet
connection. Advantage of the technology usage was that scheduling of the interviews could have been
very flexible. If I wanted to perform interviews in person, I would have to manage to fit all of them into
defined time slots when visiting Estonia. This could potentially mean more difficult access to primary
data. In addition, it conducting interviews via Skype also allowed me to phase the interviews, and thus
to be able to revise interview guide each time. If I performed interviews personally in Estonia, I would
have several interviews back to back in one day, not giving me a chance to adjust. Lastly, performing
several interviews in row in person, I could be tired and influenced by the previous interview. This was
eliminated as I had maximum one interviewee per day.
On the other hand, this setup made it harder for me to read body language and facial expressions of the
interviewee and react appropriately to them. It was difficult to recognize when the interviewee is taking
pause to think and formulate his or her answer and when they are simply waiting for another question.
This type of communication could have also negatively impacted personal connection with the
interviewees, who might open more in case we were having interviews face to face.
Interviews were recorded and transcribed in order to provide material for analysis.
Language The interviews were held in English. The original assumption was that the English language wouldn’t
be a problem as Estonians were assumed to be fluent in the language. While all of the interviewees were
indeed fluent in English, it sometimes appeared as if they had some difficulties expressing their ideas in
as much detail and as much nuance, as they would have liked to.
For the purpose of better legibility, the final quotes presented in this master’s thesis were adapted. They
were condensed and grammatically corrected to more clearly and concisely represent what the
interviewees wanted to say. This was done by the author of this thesis based on transcripts.
10
Validity and reliability of information from interviews The whole process of implementation of EHR in Estonia started twelve years ago, in 2002. This means
that memories of interviewees concerning such a long and past period may have been obscured. They
may not remember all the details anymore and their time perception may be distorted. Therefore, I tried
to compare each factual piece of information provided by interviewees with secondary documents. On
the other hand, such a long time timeframe gave the interviewees opportunity to perceive the process
from a perspective, filtering unimportant events out, and highlighting only the important ones. If the
experiences were more recent, interviewees might have felt urge to share more details, obscuring the
importance of the key events.
Interview analysis Interviews were transcribed and analyzed by the author of this thesis. In the analysis, I identified several
large usage motivation topics. Each of these topics was further supported by several smaller arguments
or examples. These are presented in chapter 0. The argumentation line in the analytical chapter 0 is built
with support of theoretical findings derived from the TAM research.
These theoretical findings are summarized in the following chapter.
11
TECHNOLOGY ACCEPTANCE MODEL In this master’s thesis, I am going to use Technology Acceptance Model (TAM) to structure analysis of
elements that led to high acceptance and usage of the Estonian national EHR. In this chapter, I am going
to outline key concepts presented in literature on Technology Acceptance Model, focusing specifically
on Technology Acceptance Model research from healthcare settings. The information provided here
should give enough background to be able to identify key elements for acceptance and usage of
Estonian EHR in the analytical chapter.
This chapter therefore represents meta-analysis of the basic TAM model article as described by Davis
(Davis, 1989) and TAM2 extended model as described by Venkatesh & Davis (Venkatesh & Davis,
2000), as well as nine other studies relevant to healthcare settings. The list of studies I go through is not
exhaustive. Some of them were generated using search functions for TAM + healthcare in CBS Library
article search engine. However, this engine does not reveal all the existing articles, so additional articles
were researched based on Holden’s and Karsh’s meta-analysis. (Holden & Karsh, 2010)
1.4 Importance of Technology Acceptance Model
The prevailing model used in the literature evaluating user acceptance and usage of technologies is
Technology Acceptance Model introduced by Fred D. Davis in 1989. (Davis, 1989) Holden and Karsch
noted that TAM is “somewhat a gold standard” and they also estimated that 10 % of IS publications
space is allocated to TAM. (Holden & Karsh, 2010, s. 159) On the same note, Ketikidis et al.
acknowledge that TAM “figures prominently among the key theoretical approaches used to understand
people’s intentions to accept various forms of information technology”. (Ketikidis, Dimitrovski,
Lazuras, & Bath, 2012, s. 125) Moores confirms that TAM is one of the most influential theories in IT
adoption and acceptance research. (Moores, 2012) The prevalence of TAM in the IS literature is also
acknowledged by Pai and Huang in their study from 2011. (Pai & Huang, 2011)
Originally, the TAM was developed by Davis to assess IBM employees’ acceptance of new software.
IBM employees may be fundamentally different to healthcare professionals. Firstly, they are working in
a corporate world, which is, compared to healthcare world tied by less regulations and more open to
innovation, and secondly, IBM as such is a technology company, so it is possible to argue that its
employees have more positive approach to technology than healthcare professionals. Therefore, many
12
scholars worried it might not be appropriate to apply the TAM developed on managers, who tend to be
early adopters, to professional workers, such as healthcare providers. (Liang, Xue, & Byrd, 2003)
These worries didn’t prove to be justified, as after research, it is also considered that TAM is fitting to
healthcare settings. (Holden & Karsh, 2010) For example, Holden and Karsh (2010) performed a meta-
analysis of over 20 studies focusing on clinicians using health IT applications for patient care. They
concluded that many of the relationships specified in TAM were repeatedly validated in healthcare
settings and fairly large proportions of variance explained in the dependent variable. Van Shaik et al.
found that original TAM model significantly predicted acceptance of a prototype system for postural
assessment by physiologists. (Van Schaik, Bettany-Saltikov, & Warren, 2002) Chau and Hu found out
that TAM is more appropriate model to understand physicians’ acceptance of technology compared to
the theory of planned behavior (Chau & Hu, 2001) Yi et al. found that perceived usefulness (PU), a core
TAM variable, played an important role in predicting PDA usage intentions by healthcare professionals.
(Yi, Jackson, Park, & Probst, 2006)
1.5 Original Technology Acceptance Model
Fred D. Davis introduced Technology Acceptance Model in 1989 in the research paper called
“Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology”.
In his basic model, Davis starts with the assumption that actual use of the technology, defined as “one
specific behavior of interest performed by individuals with regard to some information technology
system” (Holden & Karsh, 2010) is influenced directly by behavioral intention to use, which he also
calls “acceptance”. Behavioral intention is defined as “an individual’s motivation or willingness to exert
effort to perform the targeted behavior”. (Holden & Karsh, 2010) Behavioral intention, and not the
actual use, is used in the TAM.
The acceptance is influenced by attitude and by perceived usefulness (PU). Attitude is defined as
“individual’s evaluative judgment of the target behavior on some dimensions (e.g. good/bad,
harmful/beneficial, pleasant/unpleasant)”. (Holden & Karsh, 2010)
Attitude is influenced by the two key TAM variables – perceived usefulness, defined as “an individual’s
perception that using an IT system will enhance job performance” and perceived ease of use, defined as
“an individual’s perception that using an IT system will be free of effort”. (Holden & Karsh, 2010)
13
Perceived ease of use also directly influences perceived usefulness. The model is illustrated in the
following figure.
Figure 2: TAM. Source: Holden & Karsh, 2010
1.5.1 Underlying research
TAM didn’t become widely used concept in the IS by accident. It’s core concept and relevancy of core
concepts of perceived usefulness and perceived used are supported by both, theoretical findings and
academic research.
The importance of perceived ease of use and perceived usefulness, that are the corner stones of the
TAM, is supported by several other researchers and studies, that identify similar concepts having
influence on human behavior.
The first research is the self-efficacy theory by Bandura (Bandura, 1982). Bandura concludes that “in
any given instance, behavior would be best predicted by considering both self-efficacy and outcome
beliefs”. (Davis, 1989, s. 329) Bandura’s self-efficacy is similar to perceived ease of use and the
outcome judgment is similar to perceived usefulness.
The next one is the cost-benefit paradigm from behavioral decision theory by Beach and Mitchell
(Beach & Mitchell, 1978) and Johnson and Payne. (Johnson & Payne, 1985) Cost-benefit paradigm
explains how people choose among different options by comparing the effort required (similar to
perceived ease of use) and the quality (accuracy) of the resulting decision (similar to perceived
usefulness).
In research of adoption of innovations, Tornatzky and Klein (Tornatzky & Klein, 1982) found out that
compatibility, relative advantage and complexity have significant relationships across a broad range of
innovation types. While compatibility and relative advantage have been used broadly and inconsistently
in the literature, it is difficult to define them more exactly, complexity is defined by Rogers and
Shoemaker (Rogers & Shoemaker, 1971) as “the degree to which an innovation is perceived as
14
relatively difficult to understand and use”. (Davis, 1989, s. 322) This definition of complexity is similar
to definition of perceived ease of use.
Swanson (Swanson, 1982) hypothesized that users select and use information reports based on their
attributed information quality and attributed access quality, or associated cost of access, trade-off. Based
on their load factors and definitions in Swanson’s studies, these correspond to perceived usefulness and
perceived ease of use.
The four theories mentioned above support Davis’ TAM built on perceived usefulness and perceived
use. In addition to this theoretical support, TAM was also confirmed in further research. King and He
performed a meta-analysis of 88 TAM studies including more than 12 000 observations and concluded
that TAM measures - perceived ease of use, perceived usefulness and behavioral intention to use a
system are highly reliable and may be used in a variety of contexts. (King & He, 2006) However, they
also noted that the correlations reported, while strong, were rather variable, so further research for
moderator or external variables is recommended. This gave birth to multiple TAM extensions and
variations that will be discussed further in this chapter.
1.6 Technology Acceptance Model extensions
Since 1989, a large number of authors revised and extended the model. Several of these extensions took
on importance and were confirmed to improve accuracy of the model in explaining the user acceptance
of different technologies. In this section, I will have a look at the most prominent extension of TAM,
TAM2 proposed by Venkatesh and Davis in 2000 and several extensions applied to healthcare settings
to identify variables that were proven to have positive effect on user acceptance in healthcare settings.
1.6.1 TAM2
In 2000, Venkatesh and Davis in their study called “A Theoretical Extension of the Technology
Acceptance Model: Four Longitudinal Field Studies” introduced their revised model called TAM2. In
TAM2, perceived usefulness and perceived ease of use still play an important role, but attitude is
removed from the model, so perceived usefulness and perceived ease of use influence the acceptance
directly.
15
Figure 3: TAM2. Source: Holden & Karsch, 2010
Furthermore, TAM2 adds more factors influencing perceived usefulness. These new indirect factors are
image, job relevance, output quality, results demonstrability and subjective norm. Out of the new
factors, subjective norm is the most influential one, as it is the only one from the new factors that not
only influences perceived usefulness, but also behavioral intention to use, or the acceptance, directly.
1.6.2 Technology Acceptance Model extensions in healthcare setting
It has already been established at the beginning of the theory chapter that TAM has relevancy in
healthcare settings as well as other settings. However, researchers also proposed many other variables
that should extend the TAM in order for it to become more accurate and fit healthcare settings further.
Both Ketikidis et al. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012) and Holden & Karsh (Holden &
Karsh, 2010) arrived to conclusions calling for extending TAM for healthcare settings to achieve
standardization, better reporting data, exploration of additional motivations and relationships that might
even increase TAM’s ability to explain variance in the acceptance.
Pai and Huang, 2011: Adding quality variables into the model In 2010, Pai and Huang tested a model including perceived usefulness and perceived ease of use and
their effect on intention to use an information system among healthcare professionals in Taiwan. They
also tested effect of information quality on perceived usefulness, service quality on perceived usefulness
and perceived ease of use and system quality on perceived ease of use. (Pai & Huang, 2011)
16
Figure 4: Pai & Huang's research model. Source: author
In the research findings, they conclude that all the relations as indicated in the Figure 4 in their proposed
model are confirmed by their research:
• Information quality positively affects perceived usefulness,
• Service quality positively influences users’ perceived usefulness as well as perceived ease of
use,
• System quality positively affects perceived ease of use,
• Perceived ease of use and perceived usefulness significantly affect users’ intention to use,
• Perceived ease of use has significant positive impact on perceived usefulness. (Pai & Huang,
2011)
Moores, 2012: Mandatory use of systems Moores is testing effect of independent variables on three dependent variables – actual system use,
attitude towards using the system and compatibility. He considers actual system use and attitude
towards using as poor surrogates for acceptance of the system, and therefore defines compatibility with
the way end-users like to work as a better measure. (Moores, 2012)
Apart from using perceived usefulness and perceived ease of use, Moores defines two antecedent
variables to those constructs. Information quality means information accuracy, format being easy to
understand, and timeliness of information and completeness of content of information. Enabling factors
are computing support and self-efficacy, meaning the extent to which the system user believes they have
the capability to use the system. (Moores, 2012)
17
Figure 5: Moores' research model and conclusions. Source: author
After his analysis, he concluded that information quality positively influences perceived usefulness, but
does not influence perceived ease of use, and that enabling factors positively influences perceived ease
of use, but does not influence perceived usefulness. He also didn’t find any influential effects of
perceived ease of use on perceived usefulness. Finally, he concluded that perceived ease of use and
perceived usefulness have positive influence on attitude towards using and compatibility, but he didn’t
find a proof that these two elements affect actual use. (Moores, 2012)
Furthermore, Moores makes an important statement: “For a system that is mandatory the level of system
use was not significantly determined by attitude towards using, compatibility, perceived usefulness or
ease of use.” (Moores, 2012)
As usage of healthcare IT systems is very often bound to be mandatory, either on institutional level in
hospitals, or even national levels by law, he concludes that compatibility as the measure of IT
acceptance can better resolve the problem of how to overcome resistance towards the adoption of IT.
Focus on increasing compatibility in order to increase IT acceptance can, according to Moores, better
avoid low levels of use, feelings of frustrations, and potential aggressive resistance to the system.
(Moores, 2012)
Liang, Xue and Byrd, 2003: Personal innovativeness along side perceived usefulness and perceived ease of use One of the extensions was proposed by Liang, Xue and Byrd in their research of acceptance of PDA
technologies by healthcare providers. Their model is in the Figure 6. They leave out both attitude and
behavioral intention as direct determinants of actual use, and only focus on effects on actual use.
18
Figure 6: Liang, Xue and Byrd's research model. Source: author
In addition, they extend users’ beliefs: perceived ease of use and perceived usefulness, with a third
element, users’ trait: personal innovativeness. They also introduce new external variables: support as
antecedent of perceived ease of use, and job relevance and compatibility as antecedents of perceived
usefulness.
Liang, Xue and Byrd (2003) proposed an extended Technology Acceptance Model to predict actual
PDA usage by healthcare professionals. They found out that perceived usefulness and perceived ease of
use determined the PDA usage. They also found out that personal innovativeness influences usage
directly and indirectly via perceived ease of use. Perceived usefulness mediates the effects of job
relevance, compatibility, and perceived ease of use. Support affects usage through perceived ease of use.
Tung, Chang, Chou, 2008: Adding negative variables to the model In 2008, Tung, Chang and Chou proposed a hybrid Technology Acceptance Model and tested its
validity on a sample of Taiwanese hospital nurses. Their model included negative effect of perceived
financial cost on behavioral intention to use, and positive effect of compatibility, perceived usefulness,
perceived ease of use and trust on behavioral intention to use. In addition, they also proposed that
compatibility, perceived ease of use and trust affect perceived usefulness and that perceived ease of use
affects trust. (Tung, Chang, & Chou, 2008)
19
Figure 7: Tung, Chang and Chou model. Source: author
Liu & Ma, 2006: Perceived system performance replacing perceived ease of use The research paper of Liu & Ma from 2006 focused on examining user behavioral intention to use
electronic medical record by healthcare workers using questionnaires method. They tested model
presented in Figure 8. Their study concludes that perceived ease of use and perceived usefulness are
affecting behavioral intention to use. However, they also found out that perceived system performance
positively influenced perceived ease of use and improved understanding of perceived ease of use.
Nonetheless, in the case perceived system performance is in the picture; influence of perceived use on
behavioral intention becomes insignificant. Therefore, they conclude that it is perceived system
performance that is more direct predictor of behavioral intention than perceived ease of use. (Liu & Ma,
2006)
Figure 8: Liu & Ma's model. Source: Author
20
Wu, Wang & Lin, 2006: Effects of compatibility and self-efficacy on perceived usefulness and perceived ease of use In 2006, Wu, Wang & Lin published a research article that presented a revised Technology Acceptance
Model explaining intention to use a mobile healthcare system (MHS) by healthcare professionals.
Their research supported perceived usefulness and perceived ease of use influencing behavioral
intention to use. In addition, they also found evidence supporting compatibility and MHS self-efficacy
having positive influence on both, perceived usefulness and perceived ease of use. Moreover, they
found that compatibility influenced positively both MHS self-efficacy and behavioral intention to use.
Furthermore, they found out that compatibility has strongest total effect on the intention to use. (Wu,
Wang, & Lin, 2007)
Their research did not find sufficient evidence that technical support and training would influence
perceived ease of use and perceived usefulness; however, the evidence supported it positively
influencing MHS self-efficacy. (Wu, Wang, & Lin, 2007)
Figure 9: Wu, Wang & Lin's model. Source: Author
Paré, Sicotte & Jacques, 2006: Introducing psychological ownership In 2006, Paré, Sicotte and Jacques researched acceptance of clinical information systems by physicians.
They enrich the TAM construct with psychological ownership. While they proved the relationships
indicated in Figure 10, their research also indicated much stronger relationship between perceived
usefulness and attitude to use than between perceived ease of use and attitude to use. They found out
that feelings of ownership positively affect perceived ease of use and perceived usefulness. However,
compared to the model proposed in Figure 10, they found out that it is only communication and overall
responsibility that are important antecedents of psychological ownership, while hands-on activities did
21
not have significant effect on the development of feelings of ownership. (Paré, Sicotte, & Jacques,
2006)
Figure 10: Paré, Sicotte, Jacques TAM extension model. Source: Author
Chau & Hu, 2002: Highlighting effect of compatibility on perceived usefulness Chau and Hu published a research paper about acceptance of telemedicine by individual healthcare
professionals in 2002. Their study highlighted importance of perceived usefulness, and only limited
effect of perceived ease of use on attitude. They also found out that compatibility was a significant
determinant of perceived usefulness. The concept of peer influence they tested seemed to have no effect
on the attitude or behavioral intention to use. (Chau & Hu, 2002)
Figure 11: Chau and Hu's TAM extension model. Source: Author
22
Ketikidis et al., 2012: Effect of subjective norm on behavioral intention to use In their study, Ketikidis et al. tested updated TAM2 model. However, the specific model they tested
seems to be quite far from TAM2. In their hypothesis, they claim to focus on testing perceived
usefulness and perceived ease of use and their ability to predict healthcare professionals’ intentions to
use HIT and subjective norm, descriptive norm, relevance and computer anxiety to have influence on
usage intentions over and above the effects of perceived usefulness and perceived use. The author’s idea
of their model is depicted in Figure 12.
Results of their study didn’t confirm perceived usefulness’ role in the intention to use HIT, as they only
found proof for perceived ease of use to have positive effect on intention to use. In addition, they
confirmed relevancy of job relevance and subjective norm for intention to use HIT in healthcare
professionals. Computer anxiety and descriptive norms didn’t play a significant role in predicting
healthcare professionals’ intention to use. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012)
Figure 12: Ketikidis et al. model. Source: Author
1.7 TAM theory application in this thesis
It is clear that there is many different variations of TAM used both inside and outside the healthcare
settings and that some of the variables have been both confirmed and found insignificant by different
studies. In this section, I am going to summarize main findings from the previous paragraphs to identify
most prominent variables in the existing TAM research. These variables should help to analyze
interviews and formulate conclusions based on the analytical research in done on the case of Estonian
EHR.
23
1.7.1 Level of use vs. intention to use vs. attitude to a system
The most visible difference between different model variations is in the way they define “acceptance” or
rather, in the dependent variable that should reflect the acceptance of the technology by users. While in
the original TAM model, Davis defined three variables indicating the acceptance of the technology and
used them in the model:
- use of technology - “specific behavior of interest performed by individuals with regard to some
information technology system” (Holden & Karsh, 2010),
- behavioral intention to use (which he calls “acceptance”) - “an individual’s motivation or
willingness to exert effort to perform the targeted behavior” (Holden & Karsh, 2010)
- attitude to the technology - “individual’s evaluative judgment of the target behavior on some
dimensions (e.g. good/bad, harmful/beneficial, pleasant/unpleasant)”. (Holden & Karsh, 2010).
Davis claimed that actual use is influenced by behavioral intention and behavioral intention is
influenced by attitude and perceived usefulness. (Davis, 1989)
However, following TAM extensions often leave one of those variables out of the model. For example,
Venkatesh & Davis in TAM2 (Venkatesh & Davis, 2000) remove attitude from the dependent variables
and only keeps behavioral intention to use and level of use. Pai and Huang (Pai & Huang, 2011) and
Tung et al. (Tung, Chang, & Chou, 2008) only study effects of independent variables on the intention to
use the system. Liang et al. (Liang, Xue, & Byrd, 2003) leave both. attitude and behavioral intention to
use out and only study effects on actual use levels.
Despite that the results on independent variables does not seem to be influenced by this in the eleven
studies used, Moores makes an important statement: “For a system that is mandatory the level of system
use was not significantly determined by attitude towards using, compatibility, perceived usefulness or
ease of use.”(Moores, 2012) Moores concluded that compatibility as the measure of IT acceptance can
better resolve the problem of how to overcome resistance towards the adoption of IT. It also seems that
in the case of mandatory systems, TAM independent variables such as perceived ease of use and
perceived usefulness may explain “attitude” to the system, while they cannot explain level of use. From
that perspective, I will be considering “attitude” as well as level of use and intention to use as parts of
the TAM model in this thesis.
24
1.7.2 TAM variables summary
In Figure 13, I have summarized all the different variables presented in the studied research on
Technology Acceptance Model. Each tick in the Figure 13 represents a variable confirmed in the
respective research. In this part of theoretical chapter, I am going to describe the variables that were
repeatedly confirmed in the TAM research as they are the ones most probably appearing as significant
also in the case of Estonian EHR, so their detailed knowledge will help analyzing the case in hand.
Figure 13: Occurrences of different variables in different TAM variations used in this thesis. Source: Author
Perceived usefulness and perceived ease of use The most often confirmed was effect of perceived usefulness and ease of use. Each variable was
confirmed in 10 out of 11 studies selected for theoretical research for this master’s thesis. In the overall
research, scientists have often argued about relative importance of perceived ease of use and perceived
usefulness. For example, Davis found out that usefulness was more strongly correlated with usage than
ease of use. (Davis, 1989) Furthermore, importance of perceived usefulness has been consistently
supported by number of studies, results supporting perceived ease of use were less consistent as some
studies failed to show the link between perceived ease of use and usage.
Also Holden and Karsch (2010) found the relationship between perceived usefulness and intention to
use or the actual use of health IT significant. They also concluded that perceived ease of use is not that
25
likely to affect acceptance, but it appears to correlate with usefulness. It thus seems that an IT system
that is perceived as difficult to use cannot possibly be perceived as useful.
However, for the purpose of this thesis, based on the eleven studies research, it seems that both
perceived ease of use and perceived usefulness play significant roles in the healthcare settings and they
are expected to play a role in acceptance and usage of Estonian EHR system.
Perceived usefulness is defined as “an individual’s perception that using an IT system will enhance job
performance”. It means that users will positively reply to questions such as: (1) ‘the technology’ enables
me to accomplish tasks more quickly, (2) using ‘the technology’ increases my productivity, (3) using
‘the technology’ improves my job performance, (4) using ‘the technology’ enhances my effectiveness
on the job, (5) using ‘the technology’ makes it easier to do my job, and (6) overall, I find ‘the
technology’ useful in my job. (Davis, 1989)
Perceived ease of use is defined as “an individual’s perception that using an IT system will be free of
effort”. (Holden & Karsh, 2010) As he did for perceived usefulness, based on extensive meta-research
of concepts similar to perceived ease of use, Davis defined 6 items that load concept of perceived ease
of use: (1) learning to operate ‘the technology’ is easy to me, (2) I find it easy to get ‘the technology’ to
do what I want it to do, (3) ‘the technology’ is rigid and inflexible to interact with, (4) my interaction
with ‘the technology’ is easy for me to understand, (5) I find it takes a lot of effort to become skillful at
use ‘the technology’ and (6) overall, I find ‘the technology’ easy to use. (Davis, 1989)
Compatibility Out of the eleven studies researched for the purpose of this thesis, four confirmed importance of
compatibility for acceptance and levels of use of the system.
Compatibility is one of the key variables in diffusion of innovation theory. Rogers first described it in
the book Diffusion of Innovations as one of five major constructs that determine innovation acceptance.
Innovation diffusion theory greatly complements Technology Acceptance Model theory and these two
theories have been often combined together to improve predictability of TAM. Diverse studies have
shown that compatibility may be a significant determinant of perceived ease of use.
Compatibility is defined as “the degree to which an IT is perceived as being consistent with the existing
values, needs and past experiences of potential adopters”. (Liang, Xue, & Byrd, 2003, s. 377) Or in
other words: “Compatibility is the extent to which the value of the innovation, its experience in the past,
26
and users’ needs are consistent with each other.” (Tung, Chang, & Chou, 2008, s. 326) It means that if
users think that using of an IT will not affect their current working style, or will only require
insignificant changes, they are more open to use that IT. (Chau & Hu, 2002)
Furthermore, also Moores (Moores, 2012) concluded that to overcome resistance towards the system,
the system must be designed to fit with current work practices.
Support and service quality Three studies confirmed that support or service quality connected to the system have important effect on
either perceived ease of use or perceived usefulness of the system. One study showed that trainings and
support positively influence self-efficacy, which in turn positively affects perceived ease of use.
Support is defined as “objective factors, ‘out there’ in the environment, that several judges or observers
can agree will make an act easy to do.” (Liang, Xue, & Byrd, 2003, s. 377) Support includes training,
consulting, IT support, hardware and software availability etc. This kind of support is likely to increase
users’ confidence in using the technology in question.
Pai & Huang (Pai & Huang, 2011) also highlighted importance and ability of service quality that goes
together with a system to improve its perceived ease of use and usefulness. The quality service is
considered to be on time, professional and personalized.
Wu et al. studied influence of technical support and training on perceived ease of use. They found no
significant influence of training and technical support on perceived ease of use, however they found out
that these elements have positive effect on self-efficacy. (Wu, Wang, & Lin, 2007)
The training needs to be built to match existing levels of knowledge and it was suggested that also
general computer training, not only system-specific one, improved self-efficacy. However, in their
discussion, Wu et al. concluded that as the systems are designed in a more user-friendly way,
importance of training and technical support decreases. (Wu, Wang, & Lin, 2007)
Job relevance Job relevance was found significant variable in several TAM extensions, such as in TAM2 by
Venkatesh & Davis (Venkatesh & Davis, 2000), Liang et al. (Liang, Xue, & Byrd, 2003) and Ketikidis
et al. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012). Job relevance is defined as “a user’s perception
regarding the degree to which the target IT is applicable to his or her job”. (Liang, Xue, & Byrd, 2003)
27
Job relevance means that IT users compare important work goals and the consequences of using it.
Based on this comparison, they make a judgment about whether the technology is useful or not.
Assumption that job relevance is influencing IT usage is based on motivation theory, action theory and
behavioral decision theory, which state that human behavior is driven by a mental representation linking
high-level goals to specific actions. In this logic, IT users compare important work goals and the
consequences of using it. Based on this comparison, they make a judgment about whether the
technology is useful or not. (Liang, Xue, & Byrd, 2003)
Also Ketikidis et al. in their conclusions confirmed that job relevance has an effect on intention to use a
system by healthcare professionals. They concluded that the more healthcare professionals know about
the applicability of the system on their daily work routines, the more likely they are to accept those
systems. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012)
Information and output quality In his study, Moores proved that information quality significantly positively affects perceived
usefulness of the system and thus plays a role in TAM. (Moores, 2012) Information quality was defined
by Moores to be based on accuracy, content, format and timeliness of the information provided by the
system. (Moores, 2012) This means, to achieve high information quality, the system must be free of
errors, provide information needed for the user to finish their work, in an easy to read format and at the
time needed. (Moores, 2012)
Information quality was assumed to positively affect perceived usefulness of a system also by Pai &
Huang. (Pai & Huang, 2011) According to them, information quality reflected in ability of the system to
provide correct information and knowledge.
Subjective norm Subjective norm was found to have positive effects in terms of TAM research in TAM2 by Venkatesh &
Davis (Venkatesh & Davis, 2000) and by Ketikidis et al. (Ketikidis, Dimitrovski, Lazuras, & Bath,
2012) Subjective norm is an indicator of social influence. It is theorized in the TAM2 that subjective
norm is an effect that opinions of important others, such as colleagues, family, friends etc. have on
intention to use the technology. In TAM2, subjective form is deemed to influence perceived usefulness.
(Ketikidis, Dimitrovski, Lazuras, & Bath, 2012) This concept is different from a descriptive norm,
which also evaluates social influence on usage and intention to use, but this time rather then opinions of
28
others, it is behaviors of others (i.e. if they do or do not use the system) that would play a role in the
acceptance of the systems. Ketikidis et al. studied effects of descriptive norms, but didn’t find any
significant impact of descriptive norms on intention to use a system.
Self-efficacy and personal innovativeness Self-efficacy and personal innovativeness were each confirmed by one of the studies research, by
Moores (Moores, 2012) and by Liang et al. (Liang, Xue, & Byrd, 2003). However, the concepts are
overlapping, so they were grouped into one category for the purpose of this thesis.
Personal innovativeness is defined as “the willingness of an individual to try out any information
technology.” It is also considered to be relatively stable across time and environment. (Liang, Xue, &
Byrd, 2003) It is assumed that personal innovativeness has a strong relationship with computer self-
efficacy, which is defined as the judgment of ones capability to use an IT, which determines individuals’
IT using behavior. It is also stated that there will be a relationship between computer self-efficacy and
perceived ease of use.
In summary, it can be said that the more users consider themselves capable of using an IT, the higher
their degree of personal innovativeness, which in turn has a positive effect on perceived ease of use of a
technology.
Self-efficacy reflects “the extent to which the system user believes they have capability to use the
system. “ (Moores, 2012, s. 509) This can be measured by comfort, ease or confidence using the system
without external help.
It is however interesting, that also Ketikidis et al. studies effects of computer anxiety on intention to use
a technology in healthcare setting, but their study didn’t manage to find significant negative connection
between computer anxiety and intention to use a system. (Ketikidis, Dimitrovski, Lazuras, & Bath,
2012)
In this theoretical chapter, I have analyzed eleven TAM research articles in order to identify key trends
and variables used in the TAM research in healthcare settings. Information from this chapter are further
used in the analytical chapters in order to identify specific activities and elements that lead to high usage
29
of the Estonian EHR system by healthcare professionals. However, I will also focus on activities and
elements that helped improve attitude of healthcare professionals towards the system.
30
ANALYSIS – ESTONIAN EHR DESCRIPTION The analysis is divided into two chapters. In this chapter, I focus on describing the Estonian Electronic
Health Record system and its implementation process based mostly on secondary data sources, such as
reports, presentations and online presentations. This chapter should allow the reader to form an
overview about the scope and the implementation process of the system, allowing him to better navigate
through the following chapter.
In the following chapter, I focus on analyzing information collected during the interviews to identify the
main conditions and activities that led to a high acceptance of the system by healthcare professionals in
Estonia.
1.8 EHR in Estonian: system features, goals and vision
“The Estonian Health Information System is a globally unique e-health solution as it encompasses the
whole country, registers virtually all residents’ medical history from birth to death, and is based on the
comprehensive state-developed basic IT infrastructure.” (Tiik & Ross, 2010)
The Estonian Health Information System includes four e-health projects: Electronic Healthcare Record,
Digital prescription, E-registration and Digital Images. The main goals for starting the e-health projects
were: (1) decreasing the level of bureaucracy in the doctors’ work process, (2) increasing the efficiency
of the health care system, (3) making the time-critical information accessible for the attending
physician, (4) developing health care services that are more patient friendly and have higher quality.
(Estonian e-Health Foundation, 2010)
This master’s thesis focuses primarily on Electronic Health Record system. EHR registers the medical
history of Estonian citizens. Its main goal is to enable the exchange of information between doctors by
connecting IT systems for health services. (ProeHealth)
Apart from patients’ primary data, such as the contact information, insurance information, allergies,
important drug information etc., the EHR contains the following eleven categories of data: medical files,
time critical data (allergy, chronic diseases), GP and hospital visits, summary of ambulatory and
stationary cases, link to medical images, prescribed and dispensed medication, expressions of
31
will/preferences, closing medical records (opt out), names of trusties, donation of organs and overview
of logs. (ProeHealth)
Apart from storing the data, EHR provides additional services to healthcare professionals and other
stakeholders in order to simplify their way through the healthcare systems. These services include for
example:
- EHR gives the doctors possibility to see the patient’s entire health history when they need it.
- EHR provides time critical information to ambulances staff (for example if patient has any
allergies or if there are any drugs that are dangerous to the patient).
- EHR gives doctors the possibility to receive consultation from colleagues as they can exchange
patient’s medical information through the EHR system.
- EHR reduced the need to use paper forms carried by patient from one stakeholder to another; for
example, (a) the GP can send patient’s medical information through the EHR system to the
specialist who treats the patient. The patient does not have to carry any papers himself; (b) the
patient can receive medical certificates through EHR without having to deal with any additional
paperwork. These certificates can also be sent to employers or state authorities through EHR
system in electronic form. (ProeHealth)
1.9 Technical solution
The HIE (healthcare information exchange) platform that is the basis for the Estonian EHR utilizes
already existing state infrastructure such as electronic ID cards, X-Road security and communications
and existing ICT systems of healthcare providers. (ProeHealth) This means that Estonian national EHR
does not replace the in-house information system of healthcare providers - these must be updated and
modified to enable data update according to the technical specifications set by the system administrator.
(ProeHealth)
The existing in-house information systems and communication infrastructure were integrated ("linked")
by special system modules. The module for message exchange enables data exchange and
interoperability of all integrated users and delivers all messages, which conform to the standard message
type. (Estonian e-Health Foundation, 2010)
However, unfortunately in spite of a more than ten years history of digital collection of medical data by
doctors prior to the national EHR implementation, the medical data collected prior to 1st of January
32
2009 were not transferred to the EHR due to their poor quality and insufficient standardization.
(ProeHealth)
1.10 Estonian healthcare system
The Ministry of Social Affaires manages the healthcare system in Estonia. The Estonian Health
Insurance Fund is the single payor in the system. Its revenues consist of public financing (73,7 % in
2006) which comes from the compulsory solidarity-based public insurance system and social taxes, and
private expenditures (25,6 % in 2006). All Estonian citizens, including those unemployed, children, or
retired people, have equal access to healthcare services regardless of their financial contribution through
insurance or taxes. (Koppel, Leventhal, & Sedgley, 2009)
The system is further characterized by the purchaser (EHIF) and provider split (healthcare providers),
free choice of healthcare provider and a high level of autonomy. The system is built around countrywide
primary care with family doctors as the main providers, supported by ambulatory services. Specialized
care is being provided mostly in outpatient settings and advanced technology is centralized to fewer
institutions. (Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
1.11 EHR - Institutional organization
Ministry of Social Affaires is the main healthcare policy developer, including the Estonian Electronic
Health Record project. The projects are developed according to the Ministry of Finance’s guidelines,
discussed and approved by the Government, and may receive financing from the state budget. (Doupi,
Renko, Giest, Heywood , & Dumortier, 2010) In addition, the involvement of the Ministry of Social
Affaires is important, as the e-health projects involve multiple stakeholders and social development
activities. Additionally, they include other aspects such as medical standardization, ethics and
legislation. (Estonian e-Health Foundation, 2010)
Estonian Health Insurance Fund is the main funding agency of the project together with the EU
Structural Funds. Estonian e-Health Foundation is the main administrative body ensuring the
implementation of the system.
33
1.11.1 Estonian e-Health Foundation
In 2005, Ministry of Social Affaires together with other stakeholders established an administrative body
to deliver most of the e-health projects – the Estonian e-Health Foundation (EEHF). Its goal was to
“effectively manage the process of developing these projects”. (Tiik, 2012)
The main purpose of EEH is summarized as follows: “The Estonian e-Health Foundation promotes and
develops national e-solutions within the health care system – we create solutions and offer services with
the goal to assist in providing high-quality and accessible health care services. Our broader goal is to
promote the development of a patient-centered health care system that has well-informed patients.”
(Estonian e-Health Foundation)
The e-solutions that are part of the EEHF’s responsibility include three out of four HIE projects:
Electronic Health Record, Digital Image, and Digital Registration. The fourth e-solution, Digital
Prescription, was put into responsibility of Estonian Health Insurance Fund.
The founders of the EEHF were: Ministry of Social Affairs of Estonia, North Estonia Medical Centre,
Tartu University Hospital Foundation, East Tallinn Central Hospital, Estonian Hospitals Association,
The Estonian Society of Family Doctors and Union of Estonian Emergency Medical Services.
Each of these stakeholders is represented in the EEHF Board. There are 11 members of the Board –
Ministry of Social Affaires has 5 members and each of the remaining professional organizations or
healthcare providers holds 1 seat. The members are appointed by the founding organizations according
to the criteria described in the Statute of the Foundation. The Board members are appointed for a 3-year
term. (Estonian e-Health Foundation)
The evaluation and perception of EEHF in Estonia are both positive and negative. The defenders of
EEHF argument that EEHF had an accurate overview of the project’s progress and that it was staffed
with experts. (ProeHealth) Another argument in favor of the positive role of EEHF in the process is that
it unified various stakeholders to ensure compliance and cooperation in developing the four projects.
(Tiik, 2012)
However, there are also arguments that represent negative attitude towards EEHF. One of them is that
not all the stakeholders are equally represented - health professionals are only represented to a small
degree, including GPs, radiologists, doctors interested in telemedicine. Many crucial partners – Estonian
Health Insurance Fund, Estonian Medical Association, University of Tartu, and the patients – are not
actively engaged in development process and there is no clear strategy for communicating and
34
establishing the added value for each one of them. (Doupi, Renko, Giest, Heywood , & Dumortier,
2010) Mostly, it is also just professional associations and their representatives that participate in the
dialogue and in planning committees, so not all the doctors had the opportunity to be actively involved
in the communication. (Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
1.12 Timeline
The first idea to improve and extend health services for patients and citizens started in 2000 in the form
of Estonian National Health Information System (NHIS) and was introduced in the Estonian Health
Project 2015 document for the World Bank. (ProeHealth) This project was then terminated in 2002,
when the Health Division of the Ministry of Social Affairs took fully over the responsibilities for the
entire process. At the same time, in 2000, all primary care practices were obliged to procure computers
and Internet connection. (Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
The ideas of e-health and electronic health record already emerged in 2002. The purpose was to develop
a nationwide framework (database) to facilitate the exchange of digital medical documents and diffuse
health information available so far only in local databases and information systems that were not able to
communicate with each other. (Estonian e-Health Foundation, 2010)
In 2003, the Department of Health Information and Analysis was established for the practical
development of the project’s strategy. (Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
In 2004, the Estonian Health Information System Development Plan 2005-2008 was published. It
included targets such as improving interoperability of data exchange between different parties within the
healthcare system. In November 2014, the Estonian Digital Health Record Vision document was
published. It contained series of feasibility studies undertaken in the first years of the NHIS project).
(Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
In 2005, Estonian Ministry of Social Affaires obtained funding from the EU Structural Funds to develop
and launch e-health projects in 2005 – 2008. These included:
• Electronic Health Record (1,6 mil EUR),
• Digital images (0,2 mil EUR),
• Digital Registration (0,2 mil EUR) and
• Digital Prescription (0,24 mil EUR). (ProeHealth)
35
The funding in total amounted to 1,8 EUR per capita from the EU Structural Funds, but increased to 7
EUR per capita from Estonia’s own resources. (ProeHealth) It was planned to include approximately
1500 participants to interfere with the EHR system. The budget for drafting the e-health bill and
implementing the four e-health projects is 36 million EEK. (Estonian e-Health Foundation, 2008 )
As was also mentioned above, in chapter 1.11 EHR - Institutional organization, in 2005, EEHF was
founded by the Ministry of Social Affaires and other stakeholders.
In 2006, the Estonian government initiated the development process of the Electronic Health Record
System as a part of the Estonian Information Society strategy 2013. (Tiik, 2012)
In December 2007, the Estonian Parliament ratified an amendment act that established the legal basis
for implementing the e-health projects called The Health Services Organization Act and Associated Acts
Amendment Act. The goal of these accepted changes is to unify all information systems created for a
specific health care organization into one central health information system. (Estonian e-Health
Foundation)
The basic idea of this Act is to create a basis to enable the electronic processing of different medical
documents. (Doupi, Renko, Giest, Heywood , & Dumortier, 2010)
In 2008, The Health Information Act was ratified by Parliament. It specified the content of information
stored centrally and their usage by the law. (ProeHealth)
From January to June 2008, EEHF arranged Information Days at thirteen Estonian hospitals, discussing
the current status of e-health projects, information security and upcoming implementation of the
information systems in healthcare. (Estonian e-Health Foundation)
In August 2008, TNS Emor carried out two surveys that showed that doctors and patients were mostly
in favor of EHR. 97 % of doctors and 57 % of residents had heard about the EHR. 77 % doctors that had
heard of the system were in favor of it. The main benefit they considered was that all information on
patients’ exams and consultations are available from one location regardless of the health care facility
where a patient saw a doctor, whereas the main disadvantage perceived was that initially it can be too
time consuming to enter and retrieve information from the system as well as the reliability of the system
during power failure, for instance. Overall, out of those that had heard about the system 71 % were in
favor of it. (Estonian e-Health Foundation, 2008 )
36
On 17th December 2008, the national Electronic Health Record was launched and data collections
started. (Estonian e-Health Foundation, 2010)
Since 1st January 2009, all the healthcare providers were obligated to send an agreed number of
standardized documents to EHR. (ProeHealth) In October 2009, the Patient Portal was launched.
(Estonian e-Health Foundation, 2010)
By March 2010 the EHR contains over a million medical documents. For the moment the agreed
amount of data is sent to the system, containing discharge letters (inpatients and outpatients), referrals
and links to digital images. (Estonian e-Health Foundation, 2010)
In the following chapter, I focus on describing motivations and activities that led to high usage levels of
the system by healthcare professionals.
37
ANALYSIS – ADOPTION OF ESTONIAN EHR BY HEALTHCARE PROFESSIONALS
In this chapter, I am going to describe the reasons that led to the usage of the EHR system by healthcare
professional. The reasons were identified based on analysis of interviews as described in the
methodology chapter. The analysis was structured based on the TAM variables presented in the
theoretical chapter of this thesis. The aim of this chapter is to describe the case of Estonian EHR
implementation and the activities that lead to high usage of the system in the country based on TAM
theory.
Firstly, I discuss appropriateness of the TAM for this particular case. During the interviews, it occurred
that the main usage-driving factor was a law that made using the system mandatory. In the TAM
research, mandatory usage was not vastly explored; therefore, the first section of this chapter explains
why TAM is still relevant even in the case of mandatory usage. Subsequently, I focus on describing
further elements that played role in creating positive attitude towards the system and assess their fit into
the TAM theory.
1.13 Mandatory system usage and TAM
The Estonian EHR was selected as a case for this master’s thesis because it is promoted as one of the
most successful national implementation of an EHR system in terms of high levels of usage of the
system by healthcare professionals and amount of information uploaded to the system. It seems like
such a scenario could be rich in acceptance building elements, and thus would be a good case to
illustrate different variables found in TAM theory. However, during the interviews, it became apparent
that mandatory usage of the system anchored in Estonian legal system was cited as the most important
reason for the high usage number by all the interviewees. This mandatory usage of the system has not
been deeply described in the TAM theory articles selected in the theoretical research for this thesis.
When asked about their opinion, what are the main reasons why so many healthcare professionals are
using the system, all the interviewees agreed that the most important decision was that it is obligatory by
law to send medical case summaries into the system as of September 1, 2008. The law was ratified in
December 2007, which gave healthcare providers eight months to make arrangements to comply with
the law.
38
The law included these provisions:
• As of September 1, 2008, healthcare providers are obligated to forward medical data to the
national EHR,
• Only healthcare employees currently associated with patient’s treatment have the right to make
enquiries about patient’s data. Enquiries outside of treatment process are not allowed. The data
will only be issued to healthcare employees registered with Healthcare Board,
• Patients can exercise their right to set restrictions (incl. situations such as emergency situations)
of access to their health data using Patient Portal. (Estonian e-Health Foundation, 2010)
The law became a powerful motivator for usage, despite the fact that interviewee 1 from the EEHF
revealed that, especially at the beginning, there was no real penalty to healthcare providers who didn’t
follow the rule.
“Yes, there were some punishments defined, for example to take away their medical licenses.
But we never did it, because there is never enough of doctors.” (Interviewee 1)
The fact that the law was executed despite the lack of punishments may have several reasons. Firstly, it
is tendency of (most of the) citizens to follow the law even if the probability of punishment is low.
Secondly, EEHF made agreement with the Estonian Health Insurance Fund to only sign contracts with
healthcare providers who proved they have made usage agreements of the system with EEHF. While
this is not specifically a punishment, it was an effective pre-emptive measure to ensure healthcare
providers did not break the law. (Interviewee 1, EEHF)
A controversial decision It was clear from the interviews that passing this law was a controversial decision. The controversy is
apparent from following quote of an EEHF employee:
“I think that the law that says that it’s mandatory to use the system has been the first driving
force. The other one is that Estonian Health Insurance Fund, which is funding all the healthcare
services, has put it to the agreements with the healthcare service providers. But these are not
really good examples to give.”(Interviewee 1)
Other interviewees had a similar attitude to the law that made uploading the documents to the system
obligatory. Interviewee 4 from EEHF called it “the most undiplomatic decision we took” despite that,
39
he also agreed that this decision had caused the high usage of the system by healthcare professionals in
the first place.
1.13.1 What made it possible to put the mandatory usage into the law?
That the ratification of the law was a rather extreme solution was admitted only with hesitation by the
interviewees. The question therefore is, what made it possible to ratify this law in parliament, by
politicians who are traditionally very careful about passing controversial laws, as they strive to be re-
elected. There were two possible reasons for passing the law identified while researching information
for this thesis. First was a strong commitment of the Estonian politicians to e-services and second was
financing of the EHR system.
Governmental commitment to e-services Since its independence, Estonia departed on the path of fast development of a society based on
information technology. In 2000, it offered its citizens opportunity to file their tax declaration online,
and today, 94 % of declarations are done online. In 2003 electronic ID bus tickets were introduced in
Estonia. In 2005, they organized first i-voting and in 2011, 24 % of votes were cast online. (Tiik, 2011)
These activities were launched by document called the Estonian Information Policy, which was
approved in Parliament in 1998. In 2006, Electronic Health Record became part of Estonian Information
Society Strategy 2013 initiated by Estonian government. (Tiik, 2012) These facts show that there was
pretty strong and continuous commitment of the politicians to implement various e-services in the
country.
In addition, in order to enable the emergence of these various e-service, Estonia had to invest into a
nation wide infrastructure ensuring secure access to these services. The first part of the infrastructure is
a national electronic ID card that includes a certificate for authentication and a certificate of digital
signature. The card was introduced in 2002. The second part of the infrastructure is X-Road, the
Estonian countrywide data exchange platform. Usage of both is free of charge for both institutions and
citizens. (ProeHealth)
Having these two elements already in existence and free to use significantly decreased procurement
costs of the EHR system, thus decreasing political resistance as the investment was lower. In addition,
finding new additional usages for this infrastructure would improve ROI of the investment into the
infrastructure, which is also beneficial for the political image of the government.
40
EHR system financing structure In 2005, Estonian Ministry of Social Affaires obtained funding from EU Structural Funds to develop
and launch e-health projects in 2005 – 2008 for a total amount of 2.8 million €. This money had to be
spent and billed by end of 2008. The total investment into the e-health services between 2006 and 2008
was 6.3 million €. (ProeHealth) This means that the EU investment represented large proportion of the
budget. It is therefore clear that it was crucial to use the funds in compliance with the conditions and the
timeline imposed.
This pressure was also expressed by the respondent no. 5 who sat on the advisory board of EEHF:
“And I also understand the Ministry [that they pressured the HCPs to use the system], because
they have applied money from European funds and their timetable was quite strict and they had
to spend the money and show the results in very short notice.” (Interviewee 5)
It therefore seems that another reason for the political pressure and willingness to use unpopular, top-
down solutions, such as enforcing usage by law was also external pressure from the EU Structural
Funds.
1.13.2 How does mandatory usage influences TAM?
The theoretical research shows that if a system is mandatory to use, then the usage level of system is not
significantly determined by attitude, compatibility, perceived usefulness or perceived ease of use.
(Moores, 2012) Adams et al. made similar conclusions. They concluded that if a system is used on a
compulsory basis as part of person’s job description, factors such as usefulness and ease of use may
have little influence on overall levels of use, though they may influence measures such as user
satisfaction. (Adams, Nelson, & Todd, 1992)
However, as it was described in the theoretical chapter, in the original Technology Acceptance Model
(Figure 2, p. 13), the variables such as perceived ease of use and perceived usefulness influence usage
levels through attitude and behavioral intention to use (acceptance). The acceptance (intention to use) is
defined as “an individual’s motivation or willingness to exert effort to perform the targeted behavior”
(Holden & Karsh, 2010) and it is directly influenced by attitude, which is defined as “individual’s
evaluative judgment of the target behavior on some dimensions (e.g. good/bad, harmful/beneficial,
pleasant/unpleasant)”. (Holden & Karsh, 2010) This distinction between usage, intention to use and
attitude towards the system was part of the original model constructed by Davis, but abandoned in later
41
extensions, such as in TAM2 (Venkatesh & Davis, 2000), Pai & Huang (Pai & Huang, 2011), Tung et
al. (Tung, Chang, & Chou, 2008) or Liang et al. (Liang, Xue, & Byrd, 2003).
In the case of Estonian EHR, intention to use (acceptance) and usage were highly influenced by the law
that mandated the use of EHR system and to send medical data into the system. Therefore these two
dependent variables, the independent TAM variables such as perceived usefulness and perceived ease of
use, were not crucial.
However, in this thesis, it will be described how these basic TAM independent variables played a role in
building positive attitude towards the system. Before I move onto that, it is crucial to understand why
attitude is important in the case of mandatory healthcare systems.
Importance of attitude for mandatory public systems I found three main reasons that justify importance of attitude. The first reason is that public systems
require public support. This is a basic implication derived from the way these systems are being decided
about and funded – by politicians seeking public support in the next elections. Second reason why a
positive attitude is important for mandatory healthcare systems is that it minimizes attempts to find
loopholes in the law. This reason is based on the interview analysis of the Estonian EHR case. Thirdly, a
positive attitude also maximizes the usage of the system for tasks that are not mandated in the law. This
reason is based on three stages of launching new technology by Scarborough & Zimmerer.
(Scarborough & Zimmerer, 2000) All three reasons are described more in depth in the following
paragraphs.
1) Publicly funded projects require public satisfaction
In the publicly funded sectors, such as healthcare, the systems are paid from public resources and their
implementation usually derives from a political decision to invest money into them, as compared to
investing into a different initiative. If the system is vastly unpopular, the healthcare professionals and
public can create enough pressure on politicians to discontinue the system, or to dismiss or not to re-
elect those in charge of implementing it. The success of such system is therefore not only determined by
levels of usage, but also by the attitude towards the system.
2) Positive attitude minimizes trials to find loop holes in the law
In Estonia, while it is obligatory to “forward medical data to the national EHR”, it may be difficult to
create clear and specific enough definitions of what “medical data” means to make it easy to understand
and enforce the law; and when such data should be forwarded to the system.
42
In Estonia, there are two different key documents that are to be sent to the national EHR – the hospital
case summaries issued whenever a patient leaves hospital, and ambulatory case summaries, which are
issued whenever a patient’s condition for which he visited a doctor ambulatory is terminated. While it is
estimated that more than 90 % of hospital summaries are submitted to the system, this number is only
around 50 % for ambulatory summaries. (Interviewee 1,2)
Interviewees 1 and 2 agreed that the main difference between these two types of documents is that while
the hospital summaries are well defined, the ambulatory summaries are not. It is hard to determine when
does an ambulatory case ends, in some cases patients do not report when they are cured, and in the case
of chronic diseases, the case never ends, thus there is no final statement and no ambulatory case
summary.
“The ambulatory part is pretty problematic because there are some tricky regulations and the
definitions are not clear enough. And many doctors do think they do not have to send all the
information to the central system. Let’s say when they found out the person is not ill; they decide
they won’t say it. But according to the law, they have to say it.” (Interviewee 1)
“But ambulatory it is actually hard to guess the right number because there is not clear rule
when the summary should be sent. Is it after every visit or is it after the patient does not come
often, after 2 – 3 visits? Is the person well or do they just don’t come? Ambulatory visits are not
very well established as a tool where to send documents. “ (Interviewee 2)
This example well illustrates that attitude towards a system is important. If attitude is positive, then
healthcare professionals will want to forward data to the system, so they will not question the
accurateness of the definitions provided and will behave according to “spirit” of the law or the system
and they may even try to proactively understand foggy definitions. However, if attitude is negative,
healthcare professionals will not proactively try to understand foggy definitions and they may even try
to search for loopholes in the definitions in order to avoid using the system. Therefore, cultivating a
positive attitude towards the system positively influences usage of the system.
3) Positive attitude maximizes usage above the mandatory usage
In Estonia, while it is obligatory to “forward” medical data, it is not obligatory to read data from the
system. Simon et al. suggested that the benefits will only be realized when clinicians use “higher level”
functions supporting advanced clinical decision-making. (Simon, et al., 2007) So in order to realize the
full benefits of the system, it is crucial that healthcare professional access patient data in the system and
use them in their clinical work. Before existence of national EHR, they could only read data from their
43
own systems. For numerous specialists and hospital visit, that meant only accessing referral report from
a GP. With the national system, it needs to become doctors’ routine to check data in the system, for
example measuring blood sugar for a diabetes patient. Scarborough & Zimmerer define three stages of
launching new technology. They claim that first, new technology only substitutes old appliances to
perform the same tasks. In the second stage, users ascertain they can do new things with the new
technology. And in the third stage, they revolutionize way they worked to fully fit all the opportunities
of the new technology. (Scarborough & Zimmerer, 2000)
In order to realize full benefits of the system, the users need to move into the third stage. However, this
can be realized by law only with extreme difficulties (if it is even possible), as it requires changes in
process and routines, which are extremely personal and difficult to track and enforce. In addition, it is
the best if each and every user finds the ways the system fits their style. But for users to be willing to do
so, they need to have a positive attitude towards the system.
As was shown in this section, in the mandatory healthcare systems implementation, it is not only level
of usage that matters, but also the attitude. In non-mandatory usage, it may be assumed that attitude and
acceptance are identical, as positive attitude translates into intention to use. However, in the case of
mandatory healthcare systems, the law drives intention to use and usage, and they both differ from
attitude. This implies that for mandatory healthcare settings, the distinction made between attitude,
acceptance and level of use suggested in the original TAM by Davis (Davis, 1989) should be made.
By selecting this case for this master’s thesis, I intended to select an extreme case – a case that would
especially well described motivations and activities leading up to the high usage levels. (Flyvbjerg,
2006) However, it seems that in terms of TAM theory, I have unintentionally chosen a so-called critical
case. (Flyvbjerg, 2006) If the TAM model works for this case, where the motivations may be obscured
by the mandatory usage, than it might be working for all the other cases.
In the following sections, I will investigate if TAM variables appear as influencers of attitude towards
the Estonian EHR.
44
1.14 Role of perceived usefulness in Estonian EHR implementation
In the TAM theory, perceived usefulness is one of the most important elements influencing usage and
acceptance of the technology by users. In the research articles studied in theoretical part, only one study
didn’t find connection between perceived usefulness and usage, intention to use or attitude to using a
system.
For a system to be useful, TAM theory says that it needs to enhance job performance by enabling users
to accomplish their tasks more quickly, increasing productivity and effectiveness on the job and making
it easier to do the job. (Davis, 1989) Several other independent variables were found to positively
influence perceived usefulness. These are for example job relevance, output or information quality etc.
In the case of electronic health record in Estonia, perceived usefulness was reflected in two main topics.
One of them was information availability in the system, and another theme was providing services based
on the information available in the system.
1.14.1 Information availability critical for perceived usefulness
The vision and purpose of the Estonian national EHR was to facilitate an easy exchange of information
about patients between healthcare providers. It is therefore no surprise that during the interviews,
availability and quality of information appeared as a key theme influencing perceived usefulness of the
system.
The system is only perceived useful for healthcare professionals if they think they can access
information about their patients that they couldn’t access otherwise, or it if they can access it easier or
faster than without the system. This perception is expressed in number of search queries they perform in
order to access this information. The development of search queries can be seen in the Figure 14. It is
apparent that in 2009, the number of queries was extremely low and that is has increased exponentially
since then until 2013.
The feeling of interviewees was that the reasons behind the low usage at the beginning and the
subsequent growth was the lack of sufficient information in the system at the beginning. But the amount
of information was rapidly increasing thus explaining the growth thereafter. Interviewees 1 and 4
expressed it in similar terms. Interviewee 4 said:
45
“So I would say that it takes off gradually. It is there, and it is big system, so the benefits are
coming quite slowly. But now physicians have accepted it and the usage is growing very
rapidly” (Interviewee 4)
Figure 14: Development of queries to access data in EHR in Estonia between 2009-2013. Source: Artur Novek, 2013
Apart from motivating using the information in the system, interviewee 2, an IT expert from EEHF
assumed that this motivates uploading information to the system as well: “Getting information
motivates also putting the information there.”(Interviewee 2)
The case shows a potential challenge in assuring usefulness of systems based on the sharing
information. In order to be useful, these systems need to have large amount of information available, so
that when users search for something, it is there. But there will never be enough information, if they do
not start uploading the information to the system first. It becomes a vicious circle. Tension between
administrators, who claim that users need to upload information first for system to become useful, and
users, who require the system to be useful first before starting to use it, increases.
This tension was illustrated by interviewee 3 from the Ministry of Social Affaires:
“If you want to get something back, you have to provide something.” (Interviewee 3)
This conflict can be further deepened by users attempting to input information in the easiest way
possible, preferably in open text fields, on one hand, and the need of users receiving data from their
colleagues in a structured, complete and correct format on the other hand. This is how the conflict is
illustrated by the Estonian EHR case:
46
“Doctors don´t like to click too many times. It means they like to write down everything in plain
text fields. But if you have all information in plain text it´s not very standardized, not very usable
in order to use in other documents so we have to find balance also here.” (Interviewee 3)
The interviewee 1 mentioned in her interview that to break the vicious circle (or not to even enter it), a
fast roll out pushed from top down was important, because only by forcing the healthcare professionals
to upload the information, they could manage to fill in the system with data as soon as possible and thus
start bringing benefits to the users as well.
“It may have been too much of pressure from the governmental side. But if there are no laws,
it’s much harder to get that kind level of usability that we have in Estonia”. (Interviewee 1)
As was mentioned earlier, this was achieved in Estonia by passing the law that made sending medical
data into national EHR compulsory for all healthcare providers.
This gave Estonians opportunity to shorten and overcome the period when there were not data in the
system and to force healthcare professionals to use the system anyway in order to achieve critical mass
of data that eventually would make the system useful. Importance of achieving critical mass of users or
critical mass of data was also mentioned by Adams et al. in their work. They considered it the key
ingredients in implementation success. (Adams, Nelson, & Todd, 1992)
The fast, strong, top-down implementation also had another positive effect from the point of view of
interviewee 2, EEHF employee and a medical doctor:
“At the beginning, no one knows what is going on, but later on as the knowledge is there, then
the management boards [of the hospitals] will start to consider the system like a competitive
solution which will somehow limit their possibilities in the existing markets.
And many countries have failed because of that. We were lucky because we built and launched
the system really fast, and it was very strongly guided from the top down so there was political
support to make it ready and pass the legislation to make it compulsory to send medical data to
the system, which is really important.” (Interviewee 7)
This quote illustrates that without the strong, top-down push, the critical mass may have never been
achieved, because before it is achieved, healthcare providers realize they are loosing their power and
they will increase the resistance against the system.
47
One of the key issues affecting perceived usefulness of the system was amount of data available. In the
TAM theory, amount of data in the system is closely similar to information and output quality
mentioned by several authors, such as Venkatesh & Davis (Venkatesh & Davis, 2000), Pai & Huang
(Pai & Huang, 2011) or Moores. (Moores, 2012) These authors relate this variable to perceived
usefulness of the system.
Pai & Huang (Pai & Huang, 2011) define information quality as the ability of a system to provide
correct information. If there is no or very little information in the system, the system obviously does not
have this quality. This was the case of Estonian EHR at the beginning of its existence.
Moores (Moores, 2012) defined information quality by accuracy, content, format and timeliness. This
means that in order to achieve high information quality, the system must be free of the errors, provide
information needed for the users to finish their work, in an easy to read format and at the time needed.
In the case there is little information in the system, users perform lot of queries with zero results, which
they may consider as an error of the system, and they have hard time finding information they need to
finish their work.
Right after the introduction of the system, based on these definitions by Pai & Huang and Moores,
Estonian EHR was not providing quality data to healthcare professionals, which resulted in the low
levels of usage of the system to retrieve the data from the system. As time went by, the system became
filled in with data and thus healthcare professionals started to perform more and more queries.
Increasing amount of information available made the system more useful for healthcare professionals,
resulting in higher usage levels.
Estonian EHR system was able to augment the number of queries to the system by increasing perceived
usefulness over the time. This was achieved by increasing the amount of information available in a
steady pace thanks to a fast, strong, top-down implementation approach represented mainly by
legislation making the uploading the medical data into the system mandatory.
Here, the Estonian EHR case supports the importance of information quality for perceived usefulness
and levels of usage of systems as indicated in the existing research by Moores, Pai & Huang and
Venkatesh & Davis.
In addition, since the amount of information available had visible effects on accessing information from
the system (Figure 14) by healthcare professionals, a type of usage not mandated by law, I can conclude
48
that perceived usefulness probably positively influenced attitude to the system, which in turn positively
influenced acceptance and usage of it.
1.14.2 Improving perceived usefulness by introducing e-services
Apart from the amount of information available in the system, the interviews with administrators of the
national EHR in Estonia revealed that it was also specific additional functionalities enabled by this
information that improved perceived usefulness of the system. These functionalities are for example
decision-making support for disability classification, and digitalization of the process to issue medical
certificates for driver’s licenses. The interviewees called these additional functionalities “secondary
usage of EHR” or “e-services”.
Interviewee 3 from the Ministry of Social Affaires seemed satisfied with the effect that introducing such
additional services had on the perceived usefulness:
“It promotes the system well, because then people see that they can get something back.
Especially for family doctors, life became much easier, because earlier they had to go through
all this paper record to see, what is history of the patient.” (Interviewee 3)
These additional e-services introduced on the top of the data structure improved the relevance of the
system for doctors and their daily activities. This corresponds to independent TAM variable called job
relevance. It was proved by Liang et al. (Liang, Xue, & Byrd, 2003) and Ketikidis et al. (Ketikidis,
Dimitrovski, Lazuras, & Bath, 2012) in their research. Ketikidis et al. specifically concluded that the
more healthcare professionals know about the applicability of the system on their daily work routines,
the more likely they are to accept those systems. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012) The
two e-services mentioned as examples at the beginning of this section simplify specific tasks that
healthcare professionals have to perform as part of their daily routines. Providing specific functionalities
for these tasks increases the relevancy of EHR for the existing daily work of the professionals, which
increases perceived usefulness of the system.
The introduction of e-services seems to be important in the case of Estonian national EHR, because
Estonian national EHR is a document-based system. This means that queries about a patient returns a
list of summary documents that a healthcare professional needs to read through in order to get to know
the information they need. This means that primary usage to access patients’ data is rather time-
49
consuming. Interviewee 6 from the Estonian Health Insurance Fund expressed frustration over this type
of data sets:
“Doctors want practical information and that is a big difference. The system only gives back the
doctors the documents, but the doctors want information aggregated from lot of different
documents. If the system gives to the doctor lot of documents they do not have any time to read
it.” (Interviewee 6 from Estonian Health Insurance Fund)
Interviewee 3 from the Ministry of Social Affaires also highlighted that sifting through all the
documents and all the data is time consuming:
“So it depends on how much time the doctor has and how big is the problem, if the problem is
very big then of course the doctor finds this time but if you have quite minor problem there is no
time to dig into.”(Interviewee 3)
In the healthcare settings, speed is often crucial in treating patients and healthcare providers struggle
with ever increasing expectations about number of patients seen and decreasing time needed to see one
patient. Therefore, anything that takes too much time to use won’t be perceived as useful by healthcare
professional. In this context, the e-services made it easier to achieve positive impact of the EHR on
everyday actions and goals of healthcare professionals, increasing the perceived usefulness of the
system.
In the case of Estonian EHR implementation, two of the aspects highlighted in the interviews as aspects
improving acceptance of the system can be tied to the TAM’s perceived usefulness key variable. First
one, amount of information available, can be further identified with information or output quality as
defined by Ketikidis et al. (Ketikidis, Dimitrovski, Lazuras, & Bath, 2012), Pai & Huang (Pai & Huang,
2011) and Moores. (Moores, 2012) The case describes that low information quantity at the beginning of
the implementation process caused low usage levels, as the system was perceived not to provide useful
information to the healthcare professionals. Second aspect, associated e-services, can be tied to job
relevancy as defined by Liang et al. (Liang, Xue, & Byrd, 2003) and Ketikidis et al. (Ketikidis,
Dimitrovski, Lazuras, & Bath, 2012) In the minds of interviewees, the introduction of e-services
improved perceived usefulness of the system, which in turn should influence attitude, acceptance and
usage levels. These findings support general TAM claims that perceived usefulness positively
influences attitude, acceptance and usage levels of healthcare IT systems did reflect in the Estonian
EHR implementation case.
50
In the next section, I will analyze Estonian EHR implementation aspects that influenced perceived ease
of use of the system.
1.15 Role of perceived ease of use in Estonian EHR implementation
Perceived ease of use is another key variable presented in TAM theory that influences the usage of the
system by healthcare professionals by influencing attitude and acceptance of the system. From the
studies researched for this master’s thesis, only one out of eleven didn’t support influence of perceived
ease of use on either attitude, acceptance or level of usage.
System is perceived to be an easy one to use if the users think it will be free of effort to use it. (Holden
& Karsh, 2010) More specifically, it needs to be easy to learn to operate the system and it cannot take
too much effort to become skillful at using it, it needs to be easy to get it to do what the user wants and
it needs to be easy to interact with the system. (Davis, 1989) Furthermore, additional supporting
variables were found to influence perceived ease of use, such as support (Pai & Huang, 2011), (Moores,
2012) and (Liang, Xue, & Byrd, 2003), system quality (Liu & Ma, 2006) and (Pai & Huang, 2011) and
self-efficacy (Moores, 2012) and (Wu, Wang, & Lin, 2007).
Very strong evidence of perceived ease of use playing an important role in the Estonian healthcare
setting emerged in interview 3. Interviewee 3 from the Ministry of Social Affaires, when asked what she
thinks makes the e-prescription system much more appreciated and used compared to EHR answered:
“It´s easier actually. (laughs) Because in E-prescription you just have to select the drug or the
brand name and how many packages you want so for doctors it means its not so much writing or
typing. But of course this central health record it´s more complicated. You have to type more,
you have to have more time to look at the data etc.”
This quote suggests that Estonian healthcare professionals had a more positive attitude towards a system
that is less complicated (meaning it should be easier to learn to operate it and to do it what the user
wants) and faster to use (meaning it is easy to interact with it). This places perceived ease of use at the
focus of this section of analytical chapter.
In Estonia, there were two basic strategies that could be classified as improving perceived ease of use.
Firstly, local electronic medical records used in the hospitals and by doctors at the time of
51
implementation were not centrally changed. Healthcare providers did not have to exchange their
existing EMR systems; instead they only had to update them in order to be able to send medical data
into the national system. Existing technology was also used for logging into the system. In the TAM
theory, this variable was described as compatibility.
Secondly, focus was also put on improving general computer skills, but also specific EMR and EHR
skills of physicians. This corresponds to support, training and self-efficacy variables described in the
TAM theory.
These two themes are analyzed and described in more detail in the following paragraphs.
1.15.1 Compatibility with existing processes and systems
Before the implementation of national EHR in Estonia, usage of computers by healthcare professionals
was rather high. Data from 2007 presented by Doupi et al. suggest that 95 % of doctors were using some
kind of electronic medical record before implementation of national EHR. Doupi et al. reported that 95
% of doctors stored patient diagnoses and 84 % stored medications information electronically. (Doupi,
Renko, Giest, Heywood , & Dumortier, 2010) These data suggest that prior to national EHR system
implementation, healthcare providers were using their own electronic patient record systems.
As a consequence, it was decided that in Estonia, EEHF was only responsible for introducing the
national part of the system. It did not take care of the local systems. Healthcare providers were
responsible for ensuring that their own systems worked properly together with the national system.
“One of the important issues or items we should recognize that we did not change the hospitals
neither GPs workflows nor process by implementing central system but we just added connect
the central system to existing ones.” (Interviewee 4 from EEHF)
“So if the local system is well developed, it is not so hard to put the information together, give it
signature and send it to the central system.” (Interviewee 2 from EEHF)
This was a result of EEHF trying to make the new system as compatible as possible with the existing
practices. TAM researchers such as Liang et al. (Liang, Xue, & Byrd, 2003), Tung et al. (Tung, Chang,
& Chou, 2008), Wu et al. (Wu, Wang, & Lin, 2007) and Chau & Hu (Chau & Hu, 2002) identified
compatibility as one of the variables influencing perceived ease of use. Compatibility variable reflects
that if users think that usage of an IT will not affect their current working style, or will only require
52
insignificant changes, they are more open to use that IT. (Chau & Hu, 2002) Keeping existing systems
positively influenced perceived ease of use as it eliminates need to learn to operate a new system.
However, in contradiction to what the aggregated data from Doupi et al. presented above and interviews
2 and 4 suggested, interviews 1 and 5 uncovered that real state of usage of IT systems in the Estonian
healthcare may have not been so high. This would questions the level of compatibility achieved by not
replacing existing systems.
“The work processes in the hospital had to be changed. There were hospitals when the coverage
of the electronic documentation was higher than in the other ones. The biggest problem was that
the older doctors were not used to use computers daily.” (Interviewee 1 from EEHF)
“But in this moment, in 2008, […] doctors documented their work in paper form and then they
opened their computers and copy it manually. They called it “every day making of e-Health”
and it takes lot of work from the doctors.” (Interviewee 5 from EEHF)
Apparently, there is a discrepancy between statements of different interviewees. From some
interviewees, it seems that level of compatibility was not as high as there were massive changes in daily
routines of healthcare professionals, as the usage of local systems was not as widely spread as could be
understood from the interviews 2 and 4 and data presented above.
Introducing ID card login In addition, there was one specific change to processes that healthcare professionals needed to submit to
even if they kept using their existing local electronic patient records. As access to the national system
meant access to much larger quantity of data both for a particular patient, but also for more patients, it
was necessary to increase login security. Previously, healthcare professionals usually used only their
own login and password to access local systems. With more data present in the system, it was decided
that healthcare professionals needed to use their ID cards to login into the system. This meant that they
had to carry the card with them all day long. This was not accepted happily among doctors. Interviewee
4 says:
“We started to used ID cards which is a pretty big change in habits and other is procedures
what were implemented to really secure the access and have control who access what. This was
a big issue to explain why it is needed.” (Interviewee 4 from EEHF)
53
The decision needed to be communicated clearly to healthcare professionals. Interviewee 4 explained
that this explanation was much easier as there were all stakeholders’ representatives present on the
supervisory board of EEHF so they had interest to explain this issue well to their peers. The effect of
having users’ representatives seated on the EEHF board is further discussed in section 1.16.2 Role of
EEHF in building psychological ownership on the page 61.
The example of ID cards login introduction clearly shows that if a new system is not compatible with
existing procedures, it can increase negative response and attitude from the users and extra activities
need to be done in order to overcome the issue.
In the following paragraphs, further problems with local systems are discussed.
Problems with local systems It was discussed in the previous paragraphs that the intention was to keep processes unchanged after
implementation of the national system by keeping local electronic patient records in place. However,
further problems occurred as the local systems were not user friendly and updating them to
communicate with national system proved costly for healthcare providers. In her interview, interviewee
1 from EEHF expressed that lack of control of the development of local electronic patient records
hindered the usage of the system:
“So the biggest problem is the EEHF who is responsible for the central system couldn’t give the
exact recommendations to the local IT developers how to make the systems user-friendly. So
today, [one of the] the main reason why the doctors are not sending the information to the
central system is that their own systems are not good enough.” (Interviewee 1)
It is now visible that in this particular case, maximizing “compatibility” may have actually hindered
ease of use of the system. Despite the fact it eliminated the need to actually learn how to use a new
system, it kept systems that took too much effort to operate; or it wasn’t easy to get them to do what the
users wanted; or it was not easy to interact with them. Negative effects on perceived ease of use of
maximizing compatibility may have exceeded positive effect on the perceived ease of use.
Improving processes When EEHF realized the low ease of use of the existing systems and their connection to national
system, they started to identify most crucial issues and address them. Interviewee 2 from EEHF
54
indicated that EEHF is right now trying to gain more influence on how the local systems are going to
look like:
“Now we are trying to do this, putting standards how it should look like in the local system … It would be something very important I think and it would get faster acceptance actually.”
(Interviewee 2)
In order to improve perceived ease of use, EEHF introduced several small developments, such as
introducing electronic stamps for family doctors. Interviewee 4 from EEHF described that before
introduction of electronic stamps for family doctors in 2013 probably less than 15-20 % of them were
using the system. GPs complained it took too much time to send data into the national system for every
single visit. In 2013, EEHF introduced an electronic stamp family doctors can use to sign a several
documents at once at the end of the day. In the Figure 15, you can see how the number of documents
sent by GPs to the system increased after introducing the digital stamp. This shows that making a
system easier to use can significantly improve usage levels. Since in this specific example we are
talking about ambulatory summaries, which as was mentioned previously, were not clearly defined in
the law, and thus very difficult to enforce, this would suggest that ease of use improved attitude towards
the system and its acceptance, not only the usage levels.
Figure 15: Development of number of documents sent by GPs before and after introduction of electronic stamp NOTE: Numbers 1-12 represent months of 2012, number 13-15 represent January, February and March of 2013. Source: Artur
Novek
55
It seems like the original intention was to keep processes in healthcare institutions unchanged by not
introducing new system they need to use locally, however unexpected changes to local process occurred
anyway, delivering negative impact on perceived ease of use. This corresponds to findings from TAM
theory, more particularly the importance of compatibility variable.
However, further findings showed that compatibility variable may have actually hindered perceived ease
of use, as local systems were not perceived as easy to use and they were not well integrated to allow for
easy updating of information into the national system.
This particular study raised the issue that compatibility on its own may not be a good predicting variable
of perceived ease of use.
Perceived ease of use is not only determined by the system itself, but also by its users. More skilled
users may perceive a system easier to use compared to their less skilled peers. This is discussed in the
following paragraphs.
1.15.2 Effect of computer skills on perceived ease of use
TAM research shows that perceived ease of use is influenced by self-efficacy. Self-efficacy reflects “the
extent to which the system user believes they have capability to use the system. “ (Moores, 2012, s. 509)
This can be measured by comfort, ease or confidence using the system without external help. Self-
efficacy reflects both, the subjective opinion about one’s skills and objective ability to actually use the
system. The more skills and experience with technology people have, the more they believe they are
able to operate such system. The less skills and experience they have, the less they believe they can
handle the technology.
This is reflected in a quote from interview 3, where interviewee from the Ministry of Social Affaires
revealed that older doctors with less computer skills had more negative attitude to the system:
“Doctors who are a bit older and who still don´t like to use computers, they think that it´s
bullshit actually (laughs). But younger doctors who grew up with computers see it as a quite
normal part of their work. If computer skills of doctor are good then he can type and orient
much easier way in the system and earlier. It depends on skills.” (Interviewee 3)
56
Despite the data showing 100 % Estonian GPs using computers in their practices or 95 % of doctors
storing diagnoses electronically (Doupi, Renko, Giest, Heywood , & Dumortier, 2010), the perception
of administrators was that computer skills of doctors were not high:
“When we started the system, […] some doctors didn´t know anything about computers or their
computer skills were quite limited” (Interviewee 3 from Ministry of Social Affaires)
In order to improve perceived ease of use, interviewees 1, 2, 4 and 7 revealed that EEHF organized
training sessions to improve computer and EHR literacy among healthcare professionals.
Interviewee 7 estimated that around 70 % of the healthcare professionals went through the trainings and
interviewee 4 estimated 10-15 % of budget was allocated for this educational project. The relatively
high percentage of the budget allocated would indicate that a high importance was placed on their role
in the implementation process by EEHF. The trainings were voluntary and mainly family doctors went
through the centrally organized trainings, as doctors working in hospitals were trained by the hospitals.
(Interviewee 1) Trainings were not only in-class lessons, but also e-learning. There were 24 e-learning
courses including 12 courses in Estonian and 12 in Russian because there are many Russians in Estonia.
E-learning courses included guidelines and demos.
Trainings performed had several levels to fit level of knowledge of different groups. From basic
computer literacy trainings, through training on secure access and privacy and what were the user rights.
There were also training sessions to show how to build new documents, what standards and
classification to use etc. This training structure corresponds to the suggestions of Wu et al. that the
trainings need to be built to match existing levels of knowledge and it was suggested that also general
computer training, not only system-specific one, improved self-efficacy. (Wu, Wang, & Lin, 2007)
The TAM research found out that trainings have an effect on perceived ease of use. Trainings are
usually part of a larger variable called support, which usually includes trainings, user support and other
activities aimed at helping users. Support was found important for either perceived ease of use or self-
efficacy by Moores (Moores, 2012), Wu et al. (Wu, Wang, & Lin, 2007), Liang et al. (Liang, Xue, &
Byrd, 2003) and Pai & Huang (Pai & Huang, 2011). The trainings have the power to improve both, the
objective ability and the subjective idea about ones ability to operate a technology, both important for
usage of the system.
57
Personal innovativeness In addition to personal computer self-efficacy, TAM theory also defines personal innovativeness to have
an effect on acceptance of the technology. Personal innovativeness is defined as “the willingness of an
individual to try out any information technology.” It is also considered to be relatively stable across time
and environment. (Liang, Xue, & Byrd, 2003) It is assumed that personal innovativeness has a strong
relationship with computer self-efficacy, which is defined as the judgment of ones capability to use an
IT, which determines individuals’ IT using behavior. It is assumed trainings cannot change personal
innovativeness in short term.
This would be confirmed by interviewee 1 from EEHF who considered personal innovativeness to be
embedded in the national culture:
“Estonians are pretty innovative and they are not scare of failing (…) and I think that was also
one of the reasons for success of the system.” (Interviewee 1)
Altogether, the interviews indicated that self-efficacy and personal innovativeness played role in
perceived ease of use of the system. While computer usage numbers and general opinion about Estonian
nature would suggest self-efficacy and personal innovativeness were rather high, EEHF decided to
devote significant time to provide trainings to make sure that the healthcare professionals have the skills
they need to use the system, but also that their computer self-efficacy increases. These factors would
have positively influence perceived ease of use of the system, improving attitude, acceptance and usage
levels.
Perceived ease of use played a role in implementation of Estonian EHR system. EEHF attempted to
keep ease of use as high as possible by not changing local electronic patient records (EPR). This
corresponds to compatibility variable mentioned in several TAM extensions. However, further
information showed that the existing EPRs features and design and their integration to the system, as
well as unexpected changes to the system, such as need to use ID cards to login, hindered the ease of use
of the system.
Significant effort was devoted to train healthcare professionals to be able to use the system. This was
done to make sure that healthcare professionals have the skills they need to operate the system and to
58
create and search for documents in it. However, the TAM theory also shows trainings improve personal
computer self-efficacy, which improved perceived ease of use by the individuals.
In the next section, I am going to describe process organization aimed at building psychological
ownership of the system among healthcare professionals. This was highlighted as another important
aspect leading up to positive attitude, acceptance and high levels of usage of the system by healthcare
professionals.
1.16 Role of psychological ownership in Estonian EHR implementation
The last element that emerged strongly from the interviews is psychological ownership. From the
studies researched for this thesis, only Paré et al. (Paré, Sicotte, & Jacques, 2006) mentioned
psychological ownership as an important variable in the TAM. That is why psychological ownership
wasn’t further dealt with in the theoretical chapter, as it was assumed that primarily variables mentioned
repeatedly by researches would play an important role in the Estonian EHR. However, in different ways,
importance of psychological ownership and activities leading to build the ownership among the
interested parties was mentioned in most of the interviews.
Paré et al. defined psychological ownership as a state of the mind in which individuals feel as though
the target of ownership or a piece of it is “theirs”. (Paré, Sicotte, & Jacques, 2006) In their research,
Paré et al. found out that support exists for the fact that psychological ownership leads to positive or
negative orientation towards change. The individuals will promote changes that they feel ownership
towards. Finally, they proposed and confirmed that user participation had significant and positive effect
on psychological ownership, as well as communication and overall responsibility. (Paré, Sicotte, &
Jacques, 2006) In their research, they confirmed that psychological ownership positively influencers
both, perceived ease of use and perceived usefulness.
In Estonian EHR implementation case, user participation played an important role in the whole process.
This was in one or another form highlighted by all the interviewees. For example, when asked what
were the specific strategies and activities that EEHF used to get healthcare providers on board to use the
system, interviewee 1 from EEHF started describing the process of making a vision of EHR. She
described that the Ministry of Social Affaires created a work group of approximately 100 members
including different stakeholders that met several times and talked about how the system should look like
59
and how it should work. In this way, she was highlighting the process of involving all the different
stakeholders into the process in order to build their ownership towards the system’s vision.
Another example was mentioned in interview 4, where the interviewee, MD and a board member
himself, highlighted that the fact that most of the stakeholders were seated on the EEHF’s board helped
overcome resistance towards the system. This was also highlighted as important by interviewee 3 from
Ministry of Social Affaires:
“And of course EEHF has itself a board, supervisory board. And this supervisory board consists
of representatives of different interest groups - from ministry, from health insurance fund, from
hospitals, from family doctors, from ambulance - so all these stakeholders can see that their
interests are represented and of course sometimes if there is something they don´t like, they can
protest as well (laughs).” (Interviewee 3)
Interviewee 3 further explained that recently, EEHF created “Medical Advisory Committee”1 in which
around 20 doctors discuss how to improve the system, what kind of features they would like to see etc.
She said: “This is now very well perceived that we have moved this way forward.”
In addition, interviewee 7 mentioned that an important change towards perception of the system
occurred when a person with MD title was appointed CEO of EEHF in 2007:
“The first thing – most of the doctors don’t want to listen to IT guys to tell them how to change
their working processes.” (Interviewee 7)
All of the above mentioned are different ways EEHF and the Ministry of Social Affaires tried to
involved users in the implementation process. They did so because they believed it was important for
long-term success of the system:
“Because maybe alone you can move much quicker than with anybody but if you have to be
successful in longer time then you have to involve everybody.” (Interviewee 3 from Ministry of
Social Affaires)
1 Medical Advisory Board was created in 2012 by EEHF and it is created by 17 doctors-experts. The objective of the Advisory Committee is to advise the Foundation and to give evaluations from the medical aspect, mediating the proposals of treatment institutions for the development of the health information system. (Estonian e-Health Foundation, 2012)
60
All of these examples clearly illustrate the notion of user participation, which leads to psychological
ownership of the system, which in turn supports both, perceived ease of use and perceived usefulness.
In the following paragraphs, I am going to go in depth describing different strategies used to increase
user participation in the Estonian EHR, as well as some of the challenges that appeared.
1.16.1 Involving healthcare professionals
While there are more users to the EHR system, it is the healthcare professionals that are the most
important ones to involve in the process to build their psychological ownership.
Interviewee 3 from the Ministry of Social Affaires admitted that at the beginning, doctors were not
involved much, however as time went on, EEHF realized they needed to involve them. But also doctors
themselves became more interested in the system and their IT skills became more advanced so they
could have had valuable input for the process. Interviewee 3 from the Ministry of Social Affaires said:
“Doctors are always so overwhelmed [by work], so they don´t find time to think about
something which is five ten years away. And if we think about ourselves ten year back, we were
very much paper based, and you think that if you make paper based things computer based it´s
the best solution.” (Interviewee 3)
As this quote implies, finding healthcare professionals that would like to be involved in the
implementation process and had the right skills to be able to do so was difficult at the beginning.
Another challenge in involving users in Estonia might have been to actually involve all the users – it is
virtually impossible to involve every single doctor in such a process. In Estonia, the decision was taken
that different healthcare professionals’ organization will find and nominate pioneers that will represent
the specialty or the organization in the process. It was convenient as interviewee 7 from EEHF noted:
“It was convenient for us. When the organizations’ representatives accepted something, then if
something went wrong and healthcare professionals were complaining, we just could say: ‘Your
umbrella organization has accepted it after internal discussion and they agreed’” (Interviewee
7)
The agreements on different issues were searched for using mainly discussions and searching for a
compromise. Interviewee 7, a medical doctor and an important player in EEHF, confirms the importance
of these discussions to gain acceptance of the system:
61
“We had lot of round table discussions. Firstly, inside every specialty and then all specialties
together. This was very crucial process making this acceptance and we found pioneers in each
field of healthcare who were ready to join this kind of discussion and they were like advocates
inside their own specialty.” (Interviewee 7)
When negotiating different agreements, for example definitions of ambulatory case studies, there were
several rounds of discussions as described by interviewee 3 from the Ministry of Social Affaires. Firstly,
the representatives of each relevant professional group met and discussed the specific topic. Then EEHF
created consolidated document based on their discussion arguments and sent the document to each of
them for approval. Afterwards, representatives discussed the issue with their peers and sent their
comments. Based on their comments, EEHF created new version of the document and called for another
meeting. This process was repeated several times until everyone was satisfied with the final version of
the document. This process could take quite a long time. Interviewee 3 noted:
“As people are doing it besides their regular job, people are quite occupied with other
activities, we can´t set very short and very strict deadlines.” (Interviewee 3)
It is therefore visible that ensuring user participation in the implementation process was not an easy
task, and that it can prolong the whole process. However, it was a necessary task to do in order to
achieve psychological ownership of the system by Estonian doctors.
1.16.2 Role of EEHF in building psychological ownership
Interviewees 1 and 4 highlighted that EEHF played an important role in involving users in the decision-
making and implementation process. Most of the stakeholders of the EHR were seated on the
supervisory board of the organization, which made them one of those making decisions about EHR
design, rules and implementation process. At the same time, being part of the EEHF made them
accountable for the organization’s goals. And since one of the goals was implementation of the EHR
system, they were accountable for it. At the same time, they were still doctors, hospital directors and
members of doctors’ unions and professional organizations’ leaders, so they were accountable to the
healthcare professionals as well and these healthcare professionals were respecting them, because they
were ‘one of them’. This combination gave those seated on the EEHF board responsibility to search for
solutions and compromise, rather then to search for what’s wrong with the system, and to communicate
these decisions positively towards their peers. Interviewee 4 described it this way:
62
“So we in the board where I was sitting, we were fighting extensively … but we had to agree
because it was our organization as well and we couldn’t not to make the decision, we had to find
the compromise. And once we agreed, then the execution process was very
straightforward.”(Interviewee 4)
However, interviewee 5 who was also sitting on the EEHF supervisory board on behalf of one of the
healthcare organizations disagreed and when asked if EEHF did anything to make hospitals and
clinicians happier about the national system, he replied:
“No, let’s say the main message was that you have to send these messages or we cancel your
certificate or license to act as medical institution.” (Interviewee 5)
This quote implies that, despite all of the afore mentioned activities, there might still have been some
healthcare professionals that didn’t feel involved enough in the process to own the project, despite the
fact they were sitting on the EEHF board.
The quote may also indicate that conditions such as making the system mandatory by law may
negatively influence psychological ownership of the system among the users, as they may feel it is
forced to use the system from the top. In addition, another condition negatively influencing
psychological ownership may be the strict implementation timeline that may not leave sufficient time to
overcome controversial issues and agree on a compromise. Here, we find a clash between positively
influencing levels of usage by the mandatory usage provision and psychological ownership.
The second reason why the decision to establish EEHF was evaluated positively in the interviews was
that the organization had clean start. That was explained by interviewee 6 from Estonian Health
Insurance Fund:
“Decision the Minister of social affaires took (to establish EEHF, rather then charging
Estonian Health Insurance Fund with the task) was made because ten years ago, our
organization did not have very good relations with hospitals and doctors. We would like to tell
them exactly, you have to do that and you have to go there, we would not negotiate. We were too
tough for that and the Ministry and hospitals have discussed that they need a little bit softer
organization who can make cooperation with the hospitals to make all this work.”
This would have positive impact on psychological ownership as well, at least in comparison with
charging Estonian Health Insurance Fund with the implementation process. Healthcare professionals
63
would be more likely to accept the system from an organization with a clean record, rather than from an
organization they have negative experience with, and that they feel is too pushy.
In Estonia, establishing EEHF had a positive effect on psychological ownership of the system, which in
turn positively influenced perceived usefulness and perceived ease of use, which in turn positively
influence attitude, acceptance and usage level.
Despite the fact that the Estonian EHR system was enforced by law, it is possible that psychological
ownership building activities played an important role in its acceptance. In other words, even though the
users were not able to decide if they could use the system or not, they could help shape the systems user
interface/functionality etc. leading to a greater level of acceptance.
In the voluntary systems, users can ‘evaluate’ perceived ease of use and perceived usefulness by either
using or not using the system, so the administrators and developers can observe usage levels and
determine if they are going the right direction or not. In the mandatory systems, this ‘feedback’ is
missing, as no matter what, usage levels should be almost 100 %. Therefore, if the administrators want
to make sure their system as accepted, they need put extra focus on involving the users in the decision
making and evaluation processes.
1.17 Summary: Elements influencing high attitude, acceptance or usage levels of Estonian national EHR
In this chapter, I have analyzed information gathered by interviews to identify key elements that led to
high usage levels of the national EHR in Estonia and explain their importance using TAM research. I
found out that enforcing the system by law was one of the key elements behind the high usage levels.
However, activities increasing perceived usefulness and perceived ease of use also appeared across the
interviews as important elements that enabled high usage levels. Lastly, activities building
psychological ownership formed another important group identified by interviewees as increasing usage
of EHR. The conclusions are in more detail summarized in the following chapter.
64
CONCLUSIONS The case of Estonian EHR was selected to be studied in this master’s thesis based on numbers showing
high usage of the system by healthcare professionals. The assumption was that if the usage numbers are
so high, and they were achieved over such a short period of time, the whole process of implementation
must have been very smooth and thus represents a best-case practice of national EHR implementation.
Subsequently, Technology Acceptance Model was selected as an analytical tool to uncover and explain
different elements contributing to high usage numbers.
However, during the analysis, it became apparent that while the usage numbers are high, the attitude
towards the system is not as positive as one would expect from such a high usage numbers. There were
many opinion differences and problems registered and described in the interviews. It also came to light
that the e-prescription system implemented in Estonia a year after EHR is much more popular and
accepted by the healthcare professionals and public compared to the EHR system.
This case study indicates that for analytical purposes, distinction should be kept between attitude,
intention to use (acceptance) and usage level of the system. This difference has been theorized in the
original TAM model proposes by Davis in 1989, however, one or more of these dependent variables
were later left out in TAM extensions, such as in TAM2 (Venkatesh & Davis, 2000), by Pai & Huang
(Pai & Huang, 2011), Tung et al. (Tung, Chang, & Chou, 2008) or Liang et al. (Liang, Xue, & Byrd,
2003). It has been shown that while intention to use and usage levels were caused by the law that said it
was mandatory to use the system, the effect of the law on the attitude towards the system was rather
negative. Several TAM variables were identified in this case to influence attitude towards the system
positively.
It was shown that positive attitude is an important condition for long term success of public systems, as
it allows for continuous political support, limits trials to avoid using system through loop holes in the
law, and supports usage of the system beyond mandatory usage, which may be important in order to
realize full benefits of the system.
The findings of this thesis suggest that building a positive attitude may play an important role in
implementation of publicly funded, mandatory systems.
Next, elements increasing positive attitude towards the Estonian EHR were studied. They were
identified to be perceived usefulness, perceived ease of use and psychological ownership of the system.
65
Perceived usefulness In the case of Estonian EHR, perceived usefulness was largely dependent on the amount of information
available in the system. At the beginning, the amount of information available in the system was low,
and so was the usage of the system to retrieve the documents. As the amount of information in the
system grew, so did the usage for retrieval of documents, which reflected increased perceived usefulness
of the system. At the beginning, the growth of the information available in the system was enabled by
the law mandating forwarding medical data to the system.
Here, the Estonian EHR case provides support for importance of information quality for perceived
usefulness and arguably also attitude towards the system, as the increasing amount of information
increased type of usage not mandated by law. This example shows how perceived usefulness influences
usage of the system as well as attitude towards the system.
Another important factor increasing perceived usefulness defined in the analysis was introduction of so
called “secondary usage of EHR” or “e-services”. This concept corresponds to independent TAM
variable called job relevance identified by Liang et al. (Liang, Xue, & Byrd, 2003) and Ketikidis et al.
(Ketikidis, Dimitrovski, Lazuras, & Bath, 2012) in their research.
The two e-services mentioned as examples in the analytical chapter simplify specific tasks that
healthcare professionals have to perform as part of their daily routines. Providing specific functionalities
for these tasks makes relevancy of EHR higher for existing daily work of the healthcare professionals,
which increases perceived usefulness of the system. It should then further influence positive attitude
towards the system and increase usage levels as well.
This case supports perceived usefulness to have impact on attitude, and consequently, acceptance
and usage levels in the Estonian EHR case. Furthermore, output/information quality was found to
be related to the perceived usefulness of this system, which is designed to facilitate exchange of
information between users. Another variable related to perceived usefulness identified in this case
was job relevance.
Perceived ease of use In Estonia, there were two basic activities designed to support perceived ease of use of the system. First,
at the beginning, in order to maximize compatibility with existing processes, it was decided that
healthcare providers should keep their existing electronic patient records and only update them to be
66
able to send medical data to the national system. Compatibility is one of the variables defined in the
existing TAM research to have positive effect on perceived ease of use.
However, in this case, several complications appeared. It was the fact that as the security had to be
increased to protect patient data, the way healthcare professionals login into the system had to change,
which was not compatible with the existing process. Furthermore, it also appeared that existing systems
were not as common as described, they were not easy to use or they were not updated to connect to the
national system in an ease-to-use way. Consequently, this potentially hindered perceived ease of use of
the Estonian EHR. As a result, EEHF had to become more active and apply corrective actions to
improve perceived ease of use of the locally used systems.
Secondly, general computer skills and specific EHR skills seemed to play an important role in the
perceived ease of use of the system. Healthcare professionals with limited skills were not in favor of the
system, so EEHF designed an extensive training project to increase both, general computer and specific
EHR skills of all healthcare professionals. This supports the TAM statements that self-efficacy and
training are important variables related to perceived ease of use.
In this thesis, I found out that perceived ease of use played a role in influencing the attitude
towards the system; and usage of the system. The case also shows that compatibility might be
difficult to achieve when introducing a new system, and that it does not have to necessarily have
positive effect on perceived ease of use. Furthermore, this case shows that perceived ease of use
may not only be dependent on system features, but also the users’ IT skills, which reflect in IT
self-efficacy, and can be improved by trainings.
Psychological ownership Third element emerging strongly from the interviews to influence attitude towards the Estonian EHR
was psychological ownership of the system as defined by Paré et al. (Paré, Sicotte, & Jacques, 2006) In
fact, only Paré et al. found support for this variable to have influence on perceived usefulness and
perceived ease of use and consequently on attitude towards a system and its usage. However, the
Estonian EHR case provided rather frequent reference to strategies and activities leading to increasing
psychological ownership of the system by the users - healthcare professionals.
It might be possible that focusing on psychological ownership is especially important in cases where
systems are enforced by law (or similar institution). Mandatory usage as such decreases psychological
ownership of the system by users, as users might feel the system usage is pushed upon them, creating
67
perception “them – administrators” vs. “us – users”, where the system is “theirs”, not “ours”. In this
thesis, numerous ways of involving users into the designing, decision making, implementing, promoting
and improving the system were illustrated. Users’ involvement in the process is found to be one of the
key building blocks of psychological ownership. In addition, it was proposed that in case of mandatory
systems, users’ involvement is crucial to receiving feedback on the system, as the feedback cannot be
derived from usage.
In this thesis, it was discovered that psychological ownership played an important role in building
positive attitude toward the Estonian EHR system. It was also proposed that ownership-building
activities might be specifically important in cases where the system’s usage is mandated by law.
68
DISCUSSION In this chapter, I am going to outline possible theoretical and practical contributions of this master’s
thesis, as well as indicate limitations to keep in mind and suggest possible directions for further
research.
Theoretical contributions In the theoretical chapter, eleven academic articles were studied to identify most common Technology
Acceptance Model variables supported by research in the healthcare field. They are summarized in the
Figure 13 on the page 24.
This thesis shows that Technology Acceptance Model may be a relevant model for acceptance of a
mandatory technology in public healthcare settings, as long as attitude is kept in the equation. This
thesis also showed that in the case of such systems, attitude is an important variable that needs to be
considered in order to achieve long term success of the system. This presents contrasts to tendency of
other TAM extension researchers to abandon concept of attitude from their models.
In addition, the thesis supports importance of perceived usefulness and perceived ease of use in TAM.
The thesis highlights information or output quality as an important variable influencing perceived
usefulness. This is probably caused by the fact the system is aimed at exchanging information between
users. Furthermore, in this particular case, job relevance emerged as an important variable for perceived
usefulness; self-efficacy and trainings played an important role for perceived ease of use; and
compatibility was attempted to keep perceived ease of use high. However, compatibility was not
entirely preserved as certain changes had to be made to achieve security of the system. Even where the
compatibility was preserved, it had controversial effects on perceived ease of use as the preexisting
systems were not perceived as easy to use.
Psychological ownership emerged as the third important variable playing role in the Estonian EHR
implementation. The thesis proposes that it is to counteract negative effects of mandatory usage that
makes the psychological ownership activities important in this case.
Overall, based on both theoretical research and the case analysis, it seems that the original Technology
Acceptance Model proposed by Davis in 1989 that included only perceived ease of use and perceived
usefulness as independent variable is valid even for mandatory public systems. However, it seems like it
69
is not possible to define a fixed set of variables supporting either perceived ease of use, or perceived
usefulness, or being on the same level as perceived usefulness and perceived ease of use in order to
achieve more accuracy of the model. The additional variables seem to differ greatly from case to case.
In the eleven studies research in the theoretical chapter, I found thirteen different supporting variables,
each re-appearing in max 4 studies.
However, it might be possible to define the sets of supporting variables for different types of the
systems being introduced in healthcare (national vs. local, mandatory vs. voluntary, information based
vs. function based etc.). After such classification, the struggle registered throughout the existing TAM
research to “refine” the model might be more successful in finding common denominators for specific
types of systems.
Practical implications The findings presented throughout the analytical chapter might be an inspiration for countries (or
regional organizations such as EU) when creating EHR implementation strategies. Such usage is
proposed also by Van Shaik et al. (Van Schaik, Bettany-Saltikov, & Warren, 2002) or Adams et al.
(Adams, Nelson, & Todd, 1992)
It was not aim of this thesis to provide practical implications, so I will only present couple of examples
here:
1. Lack of information in the system at the beginning is key turn off for the healthcare
professional. It needs to be addressed in the plan.
2. For fast implementation, political support across political spectrum is necessary for success and
should be translated into mandatory usage of the system anchored in the law.
3. Despite the mandatory usage provisions, administrators and politicians must strive to build
psychological ownership of the system by healthcare providers.
4. It seems to be a good idea to establish an independent organization charged with implementing
and administering the system. It gives the organization needed focus and clean start.
5. It seems to be important to make key stakeholders parts of this organization.
6. The process needs to be driven by people trained in medicine in order to gain respect of other
healthcare professionals, to be able to talk to them in their own language and to understand the
process that they go through daily.
70
7. In case of obsolete electronic patient record systems used locally, it might be better to design
new, user-friendly system for the healthcare professionals to interact with the national platform
rather than insist on keeping the old system hoping to leverage compatibility variable.
Limitations Most critical limitations of this thesis were time limits imposed on Master’s thesis project and relative
lack of knowledge of the author about the case before departing on the thesis journey.
Time limit imposed led to focusing on the point of view of administrators. Despite the fact that I believe
that they are the ones to have an overview about the whole process, their answers may be biased by the
fact they are the ones responsible for the system. Conducting interviews with users might uncover more
controversies and acceptance elements.
Before the project started, I had only limited knowledge of the Estonian EHR gained from a conference
speech and couple of additional articles. This brought about complications such as relatively large
portion of interview time spent on the system description and understanding, continuous development of
research scope, as well as the fact that TAM model was selected as the analytical framework before it
was revealed the system usage was mandatory. Fortunately, deeper analysis showed that TAM is
relevant also for mandatory systems; however, this was something that could have potentially had
serious effects on the whole project. However, this ignorance also brought about positive aspects, such
as bird view on the case and ability to analyze the interviews without having preconceptions.
Further research As was suggested before, further research might include deeper dive into Estonian case to discover
views of users on the elements that motivated them to use the system.
From theoretical perspective, further research might also study other similar systems (public, national,
mandatory, information focused systems) to determine if the variables that were highlighted in this case
correspond. In such a case case an extended TAM model for this type of system might be built.
71
TABLE OF FIGURES Figure 1: Interviewees, source: author ....................................................................................................... 8
Figure 2: TAM. Source: Holden & Karsh, 2010 ...................................................................................... 13
Figure 3: TAM2. Source: Holden & Karsch, 2010 .................................................................................. 15
Figure 4: Pai & Huang's research model. Source: author ........................................................................ 16
Figure 5: Moores' research model and conclusions. Source: author ........................................................ 17
Figure 6: Liang, Xue and Byrd's research model. Source: author ............................................................ 18
Figure 7: Tung, Chang and Chou model. Source: author ......................................................................... 19
Figure 8: Liu & Ma's model. Source: Author ........................................................................................... 19
Figure 9: Wu, Wang & Lin's model. Source: Author ............................................................................... 20
Figure 10: Paré, Sicotte, Jacques TAM extension model. Source: Author .............................................. 21
Figure 11: Chau and Hu's TAM extension model. Source: Author ......................................................... 21
Figure 12: Ketikidis et al. model. Source: Author ................................................................................... 22
Figure 13: Occurrences of different variables in different TAM variations used in this thesis. Source:
Author .............................................................................................................................................. 24
Figure 14: Development of queries to access data in EHR in Estonia between 2009-2013. Source: Artur
Novek, 2013 ..................................................................................................................................... 45
Figure 15: Development of number of documents sent by GPs before and after introduction of electronic
stamp NOTE: Numbers 1-12 represent months of 2012, number 13-15 represent January, February
and March of 2013. Source: Artur Novek ........................................................................................ 54
72
BIBLIOGRAPHY Adams, D. A., Nelson, R. R., & Todd, P. A. (1992). Perceived usefulness, ease of use and usage of
information technology: a replication. MIS Quarterly , 16 (2), 227-247.
Bandura, A. (1982). Self-efficacy mechanismi in human agency. Amecian Psycholigist , 37, 122-147.
Beach, L., & Mitchell, T. (1978). A contingency model for the selection of decision strategies. Academy
of Management Review , 3, 439-449.
Berg, M. (2001). Implementing information systems in health care organizations: myths and challenges.
International Journal of Medical Informatics , 64, 143-156.
Chau, P. Y., & Hu, P. J. (2002). Examining a Model of Information Technolgoy Acceptance by
Individual Professionals: An Exploratory Study. Journal of Management Information Systems , 18 (4),
191-229.
Chau, P., & Hu, J. (2001). Information technology acceptance by individual professionals: a model
comparison approach. Decision Sciences , 32 (4), 699-719.
Cormack, D. (1991). The Research Process in Nursing. Oxford: Blackwell Scientific.
Cresswell, K. M., Worth, A., & Sheikh, A. (2012). Comparative case study investigating sociotechnical
processes of change in the context of a national electronic health record implementation. Health
Informatics Journal , 18, 251-270.
Cresswell, K. M., Worth, A., & Sheikh, A. (2012). Integration of a nationally procured electronic health
record system into user work practices. BMC Medical Informatics and Decision Making , 12-15.
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information
technolgoy. MIS Quarterly , 13 (3), 319-339.
Doupi, P., Renko, E., Giest, S., Heywood , J., & Dumortier, J. (2010). EHealth Strategies: Country
Brief - Estonia. European Commission, DG Information Society and Media, ICT for Health Unit.
European Commission: Information Society and Media.
73
Duffy, M. (1986). Quantitative and qualitative research antagonistic or complementary. Nursing and
Health Care , 8 (6), 356-357.
Ellingsen, G., & Monteiro, E. (2012). Electronic patient record development in Norway: The case for an
evolutionary strategy. Health Policy and Technology , 1, 16-21.
Emmanouilidou, M., & Burke, M. (2013). A thematic review and a policy-analysis agenda of Electronic
Health Records in the Greek National Health System . Health Policy , 109, 31-37.
Eriksson, P., & Kovalainen, A. (2008). Qualitative Methods in Business Research. SAGE Publications
Ltd.
Estonian e-Health Foundation. (n.d.). Board. Retrieved October 17, 2014, from Estonian e-Health
Foundation: http://www.e-tervis.ee/index.php/en/2012-07-22-13-35-31/board
Estonian e-Health Foundation. (2012). Estonian e-Health Foundation presentation.
Estonian e-Health Foundation. (2008 , August 8). Estonian residents are in favour of implementing the
Electronic Health Record. Retrieved October 17, 2014, from Estonian e-Health Foundation:
http://www.e-tervis.ee/index.php/en/news-and-arcticles/434-estonian-residents-are-in-favour-of-
implementing-the-electronic-health-record
Estonian e-Health Foundation. (n.d.). Information Days at hospitals. Retrieved October 17, 2014, from
Estonian e-Health Foundation: http://www.e-tervis.ee/index.php/en/news-and-arcticles/437-information-
days-at-hospitals
Estonian e-Health Foundation. (n.d.). National eHealth Projects. Retrieved October 17, 2014, from
Estonian eHealth Foundation: http://www.e-tervis.ee/index.php/en/health-information-system/services
Estonian e-Health Foundation. (n.d.). Organization. Retrieved October 17, 2014, from Estonian e-
Health Foundation: http://www.e-tervis.ee/index.php/en/2012-07-22-13-35-31/organization
Estonian e-Health Foundation. (2010, August 10). Overview of Estonian Electronic Health Record
(EHR) System. Retrieved October 17, 2014, from Estonian e-Health Foundation: http://www.e-
tervis.ee/index.php/en/news-and-arcticles/432-overview-of-estonian-electronic-health-record-ehr-
system
74
Estonian e-Health Foundation. (n.d.). The Estonian Parliament legalized eHealth projects. Retrieved
October 17, 2014, from Estonian e-Health Foundation: http://www.e-tervis.ee/index.php/en/news-and-
arcticles/436-the-estonian-parliament-legalized-ehealth-projects
Flyvbjerg, B. (2006). Five Misunderstandings About Case-Study Research. Qualitative Inquiry , 12 (2),
219-245.
Hancock, B. (1998). Trent Focus for Research and Development in Primary Health Care: An
Introduction to Qualitative Research . Trent Focus .
Holden, J. R., & Karsh, B. (2010). The technology acceptance model: its past and its future in
healthcare. Journal of Biomedical Informatics , 43, 159-189.
Johnson, E., & Payne, J. (1985). Efford and accuracy in choice. Management Science , 31, 395-414.
Ketikidis, P., Dimitrovski, T., Lazuras, L., & Bath, P. A. (2012). Acceptance of health information
technology in health professionals: An application of the revised technology acceptance model. Health
Informatics jJournal , 18 (2), 124-134.
King, W., & He, J. (2006). A meta-analysis of the technology acceptance model. Information &
Management , 43, 740-755.
Koppel, A., Leventhal, A., & Sedgley, M. (2009). Public health in Estonia 2008: An analysis of public
health operations, services and activities. Copenhagen: WHO Europe.
Liang, H., Xue, Y., & Byrd, T. A. (2003). PDA usage in healthcare professionals: testing an extended
technology acceptance model. Int. J. Mobile Communications , 1 (4), 372-389.
Liu, L., & Ma, Q. (2006). Perceived System Performance: A Test of an Extended Technology
Acceptance Model. The Data Base for Advances in Information systems , 37 (2&3), 51-59.
McAlearney, A. S., Song, P. H., Jorina, M., Hirsch, A., Kowalczyk, N., & Chisolm, D. (2010). Moving
from Good to Great in Ambulatory Electronic Health Record Implementation. Journal for Healthcare
Quality , 32 (5), 41-50.
Moores, T. T. (2012). Towards an integrated model of IT acceptance in healthcare. Decision Support
Systems , 53, 507-516.
75
Nortal. (2013, April 26). Estonia leads European rankings for e-health services adoption. Retrieved
October 27, 2014, from Nortal: http://www.nortal.com/about-us/press/estonia-leads-european-rankings-
for-e-health-service-adoption
Novek, A. (2013). Estonian e-Health. Tallin: Estonian e-Health Foundation.
Pai, F.-Y., & Huang, K.-I. (2011). Applying the Technology Acceptnace Model to the Introduction of
healthcare information systems. Technological Forecasting & Social Change , 78 (4), 650-660.
Paré, G., Sicotte, C., & Jacques, H. (2006). The Effects of Creating Psychological Ownership on
Physicians' Acceptance of Clinical Information System. Journal of the American Medical Informatics
Association , 13 (2), 197-205.
ProeHealth. (n.d.). Estonian EHR Case Study. Retrieved October 17, 2014, from ProeHealth:
http://www.pro-ehealth.eu/casestudies/ProeHealth_Case_Report_Estonia_EHR.pdf
Rogers, E., & Shoemaker, F. (1971). Communication of innovations: a cross-cultural approach. New
York: Free Press.
Sarker, S., Xiao, X., & Beaulieu, T. (2013). Qualitative Studies in Information Systems: A Critical
Review and Some Guiding Principles. MIS Quarterly , 37 (4), iii-xviii.
Scarborough, N., & Zimmerer, T. (2000). Effective small business management. New Jersey: Prentice
Hall.
Simon, S. R., Kaushal, R., Cleary, P. D., Jenter, C. A., Volk, L. A., Orav, E. J., et al. (2007). Physicians
and Electronic Health Records. ARCH INTERN MED , 167, 507-512.
Swanson, E. (1982). Measuring user attitudes in MIS research: a review. OMEGA , 10, 157-165.
Tavakoli, N., Jahanbakhsh, M., Mokhtari, H., & Tadayon, H. R. (2011). Opportunities of Electronic
Health Record Implementation in Isfahan. Procedia Computer Science , 3, 1195–1198.
Tiik, M. (2012). Access Rights and Organizational Management in Implementation of Estonian
Electronic Health Record System. Tallin: TUT Press.
Tiik, M. (2011). Thinking of the bigger picture: How Estonia designed future healthcare. Tallin:
Estonian e-Health Foundation.
76
Tiik, M., & Ross, P. (2010). Patient opportunities in the Estonian Electronic Health Record System.
Studies in health technology and informatics , 7.
Tornatzky, L., & Klein, K. (1982). Innovation characteristics and innovation adoption implementation:
A meta-analysis of findings. IEEE Transactions on Engineering Management , 29, 28-45.
Tsang, E. W. (2013). Case study methodology: causal explanation, contextualization, and theorizing.
Journal of International Management , 19, 195-202.
Tung, F.-C., Chang, S.-C., & Chou, C.-M. (2008). An extension of trust and TAM model with IDT in
the adoption of the eletronic logistics infroamtion system in HIS in the medical industry. International
Journal of Medical Informatics , 77, 324-335.
Van Schaik, P., Bettany-Saltikov, J., & Warren, J. (2002). Clinical acceptance of a low-cost portable
system for postural assessment. Behavior & Information Technology , 21, 47-57.
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the Technology Acceptnace Model:
four longitudinal field studies. Management Science , 46, 186-204.
Wu, J.-H., Wang, S.-C., & Lin, L.-M. (2007). Mobile computing acceptance factors in the healthcare
industry: A structural equation model. International Journal of Medical Informatics , 76, 66-77.
Yi, M., Jackson, J., Park, J., & Probst, J. (2006). Understanding information technology acceptance by
individual professionals: toward integrative view. Information & Management , 43, 350-363.
